Welcome to the Graphite Note documentation portal. These guides will show you how to predict, visualize, and analyze your data using machine learning with no code. Graphite Note is a powerful tool designed to democratize the power of data analysis and machine learning, making it accessible to individuals and teams of all skill levels. Whether you're a marketer looking to segment your audience, a sales team predicting lead conversions, or an operations manager forecasting product demand, Graphite Note is your go-to platform.

The platform is built with a user-friendly interface that allows you to connect your data, generate predictive models, and share your results with just a few clicks. It's not just about making predictions; it's about understanding them. With Graphite Note's built-in prescriptive analytics and data storytelling feature, you can transform complex data into meaningful narratives that drive strategic actions. This documentation will guide you through every step of the process, from setting up your account to making your first prediction. So let's dive in and start exploring the power of no-code machine learning with Graphite Note.

For every lexical terms you can check on the machine learning glossary.

The following sections cover some of the most important machine learning concepts.

In Graphite Note, every user within a team is assigned a role that determines their access level and permissions. These roles help control what users can view or modify within the platform. Roles can be managed on the Roles Administration Page, accessed by clicking the wheel icon (⚙️) in the top-right corner and selecting Roles.

Role Types

By default in Graphite Note you will have two types of predefined roles that cover the most common access needs:

• Administrator: Grants full access to read and modify all entities, including datasets, models, notebooks, users, and system settings. Administrator role cannot be deleted but serve as the foundation for managing basic access levels.

• Viewer: Provides read-only access, allowing users to view entities but not edit or modify them.

Administrators can create custom roles tailored to the specific needs of their organization. Custom roles allow for detailed control over permissions. Custom roles are created by clicking New Role button, providing a name and description, and defining permissions (e.g., Read & Modify or No Access) for each module.

With both default and customizable roles, Graphite Note provides a robust system for managing user access and ensuring data security across your team.

Role Assignment

Roles are assigned to users on the Users Page. When inviting a new user to the team, administrators select a role for them, ensuring the appropriate level of access. Existing roles can also be updated later to adapt to changing team responsibilities.

Viewing Assigned Role

Users can see their currently assigned role on the Profile Information page, accessible via the profile dropdown menu in the top-right corner of the platform. This transparency allows users to understand their permissions within the system.

You can access the Profile information setup by clicking on the small user profile icon located in the top-right corner of the interface. This icon typically displays the first letters of your name and surname (e.g., “CP” for Chris Parker). Once clicked, a dropdown menu appears where you can select Profile to view or edit your profile settings.

This page is where users can manage their personal information and customize their profile settings:

• User Code: A unique identifier for the user (e.g., a system-generated code like “e056ac433952”).

• Name: Displays the user’s name. This can be updated to reflect the preferred name.

• AI Generated Content Language: Lets the user select the language for AI-generated content (e.g., “English”). All AI generated content related to different models in Graphite note will be generate in selected language. When the language is changed, you need to rerun the model to generate content in the currently selected language.

• Select Avatar: Allows users to choose a personalized avatar color to visually represent their profile.

• Email: Shows the registered email address of the user.

• Role: Indicates the user’s role in the system (e.g., “Administrator” or “Viewer”).

• Password: Offers an option to change the user’s password with a Change Password button.

You can access the Account information page by clicking on the small wheel icon (⚙️) in the top-right corner of Graphite Note, and then selecting the Account info drop-down item. This page features your account information, including active plan information and plan usege statistics along with information on different fatures included in the plan.

Assistance

A toggle option to enable assistance from the Graphite Note Support Team. By enabling this feature, users grant the support team access to their datasets, models, and notebooks. This facilitates quicker issue resolution and support for content creation. Assistance option is enabled by default.

Current Plan

Displays the subscription plan the account is currently on. Click on the Contact Sales button if you wish to upgrade, downgrade, or discuss your subscription options.

API Token

A secure token is shown (on the account info page it is partially masked for privacy) that users can use to integrate Graphite Note with other applications or systems. It provides a unique, secure key for API access, enabling advanced integration with external systems in a two ways:

• Dataset API - enables users to easily populate their datasets by sending data directly to Graphite Note, ensuring seamless data integration.

• Prediction API - allows users to request predictions based on attributes they provide, leveraging Graphite Note’s machine learning models to generate accurate business forecasts.

Click the eye icon next to the token to view the full API token. More information about API Token usage you can find in the REST API section.

Plan usage

Gives a quick overview of your plan limits, usage, and enabled features:

• Max Data Rows in Database: Shows the total data rows allowed (e.g., 10M), how many are used (e.g., 2.28M), and how many are left (e.g., 7.72M). A progress bar helps you track usage.

• Max Number of Users: Displays the total users allowed (e.g., 10), how many are used (e.g., 1), and available slots (e.g., 9). You can invite new users with the Invite User link.

• Available Dataset Plans: Lists supported dataset types like CSV, database integrations, model data, merged datasets, and BigQuery.

• Additional Features: Confirms whether key features like AI insights, API access, white-labeled notebooks, and advanced model settings are enabled under your plan.

The Users Page in Graphite Note provides administrators with tools to manage team members, assign roles, and update user details. It can be accessed by clicking the wheel icon (⚙️) in the top-right corner and selecting Users.

User List

Displays all team members with details such as user code, name, AI content language, email, assigned role, and activation status. Administrators can manage users directly from this list.

Inviting New Users

New users can be invited by clicking on the Invite user button and entering their email address and assigning a role (e.g., Viewer, Administrator, or a custom role) from the Users Page. Once the invitation is sent, a user profile is automatically created within the system, even before the user accepts the invitation. This allows administrators to manage and edit the user’s details, such as their name, role, or preferences, immediately after the invitation is issued. The profile becomes fully active once the user accepts the invitation via email.

Editing Invited user

Once a user is invited, administrators can edit their details by clicking the gear icon under the Action column on the Users Page. This opens the Edit User Panel, where the following information can be found:

• User Code: A unique identifier for the user (non-editable).

• Name: Modify the user’s display name to ensure accuracy or reflect changes.

• AI Generated Content Language: Select the user’s preferred language for AI-generated content.

• Select Avatar: Customize the user’s profile avatar by choosing from a range of color options for better visual distinction.

• Email: Update the user’s registered email address.

• Role: Change the user’s assigned role (e.g., Viewer, Administrator, or a custom role) to adjust their access permissions.

Changes are saved by clicking the Save button, ensuring the user profile reflects updated details. This flexibility allows for seamless management of team members’ information and roles.

Welcome to Graphite Note App! To get started, or if you already have one. With an active account, you can upload datasets and create machine learning models in just minutes.

Refer to our documentation to learn more about machine learning and how to apply ML models to your data to boost your business, uncover insights, and make predictions.

The Tags Page in Graphite Note is used to manage tags that help in organizing and distinguishing datasets, models, and notebooks. Tags improve searchability and allow users to categorize resources effectively. Note that only one tag can be assigned to each dataset, model, or notebook. The Tags Page can be accessed by clicking the wheel icon (⚙️) in the top-right corner and selecting Tags.

Tag List

Displays all created tags with their name, description, color, and a preview of how the tag appears.

Creating new Tag

To create a new tag, users can define its name and description while selecting a color for visual distinction. This allows tags to be uniquely identifiable and easy to manage. Tags can also be edited or deleted using the action icons, providing flexibility for keeping the tagging system up-to-date.

Tag creation in Datasets, Models and Noteboks

Additionally, tags can also be created during the process of creating a new dataset, model, or notebook, or within the settings options of an existing dataset, model, or notebook. This ensures that tagging remains a flexible and integrated part of resource management.

Machine learning is a method that uses data to teach computers to recognize patterns and key drivers, allowing them to predict future outcomes without being explicitly programmed.

No-code machine learning is a simplified approach to machine learning that allows users to build, train, and deploy machine learning models without needing to write any code. This makes advanced data analysis accessible to non-technical users, empowering business teams to harness machine learning insights without relying on data scientists or programmers.

In no-code machine learning, platforms like Graphite Note provide intuitive interfaces where users can import data, select features, and train models through guided steps. For example, machine learning, as a method, uses data to teach computers to recognize patterns and key drivers, enabling them to predict future outcomes. In a no-code environment, this process is automated, allowing users to set up predictive models by simply uploading data and selecting key variables, all through a user-friendly, visual workflow.

By removing the complexity of coding, no-code machine learning enables organizations to leverage powerful data insights faster, supporting better business decisions and allowing companies to respond more quickly to market demands.

Analytics maturity represents an organization’s progression in leveraging data to drive insights and decisions. This journey typically follows four levels:

1. Descriptive Analytics: The foundation of analytics maturity, focused on answering “What happened?” Descriptive analytics relies on reporting and data mining to summarize past events. Most organizations begin here, gaining basic insights by understanding historical data.

2. Diagnostic Analytics: Building on descriptive insights, diagnostic analytics answers “Why did it happen?” by drilling deeper into data patterns and trends. Using techniques such as query drill-downs, diagnostic analytics provides context and explanations, helping organizations understand the causes of past events. Traditional organizations often operate within this descriptive and diagnostic phase.

3. Predictive Analytics: Moving into more advanced analytics, predictive analytics addresses “What will happen?” by utilizing machine learning and AI to forecast future outcomes. Through statistical simulations and data models, predictive analytics enables organizations to anticipate trends, customer behavior, and potential risks. Elevating to this level empowers organizations to make more proactive, data-driven decisions and gain a competitive edge.

4. Prescriptive Analytics: At the highest level of analytics maturity, prescriptive analytics answers “What should I do?” It combines machine learning, AI, and mathematical optimization to recommend actions that lead to desired outcomes. By offering actionable guidance, prescriptive analytics not only predicts future scenarios but also prescribes the best course of action, allowing organizations to optimize decisions and drive strategic growth.

While many organizations remain in the descriptive and diagnostic phases, those aiming to stay competitive and drive innovation must elevate their analytics capabilities. Graphite Note is designed to accelerate this journey, helping organizations seamlessly transition into predictive and prescriptive analytics. By embracing machine learning and AI through Graphite Note, companies can transform their data into a strategic asset, enabling proactive decision-making and unlocking new avenues for operational efficiency and business growth.

Designed for modern analytics teams and individuals, Graphite Note offers three different subscription plans to match various project preferences: Sprout, Growth and Enterprise.

To help you choose the best plan for your project, you can start with a 14 days free trial, schedule a demo, or book a meeting with our data science experts.

What is Predictive Analytics?

Predictive analytics is a form of advanced analytics that uses both new and historical data to forecast future activity, behavior, and trends. It involves applying statistical analysis techniques, analytical queries, and automated machine learning algorithms to data sets to create predictive models that place a numerical value — or score — on the likelihood of a particular event happening.

What is Prescriptive Analytics?

Prescriptive analytics is a form of advanced analytics that examines data or content to answer the question "What should be done?" or "What can we do to make 'X' happen?". It is characterized by techniques such as graph analysis, simulation, complex event processing, neural networks, recommendation engines, heuristics, and machine learning.

What is AutoML?

AutoML (Automated Machine Learning) is the process of automating the end-to-end process of applying machine learning to real-world problems. It aims to make machine learning accessible to non-experts and to improve efficiency for experts by automating tasks such as data preprocessing, feature engineering, model selection, and hyperparameter tuning.

How does Graphite Note ensure data security?

At Graphite Note, we take data security very seriously. We employ robust security measures to ensure your data is protected at all times. This includes encryption of data at rest and in transit, regular security audits, and strict access controls. .

What types of data can I use with Graphite Note?

Graphite Note is designed to work with a wide range of data types. You can import data from various sources such as CSV files, databases, data warehouses. The platform can handle structured, tabular data (like numerical and categorical data).

How can I import my data into Graphite Note?

Importing data into Graphite Note is a straightforward process. You can upload data directly from your computer, connect to a database, or your data warehouse. Our platform supports a variety of data formats, including CSV, and SQL databases.

What kind of support is available if I need help using Graphite Note?

We offer a range of support options to help you get the most out of Graphite Note. This includes a comprehensive knowledge base, video tutorials, and email support. Our dedicated support team is always ready to assist you with any questions or issues you may have.

Can I use Graphite Note if I don't have a background in data science?

Absolutely! Graphite Note is designed to be user-friendly and accessible to everyone, regardless of their technical background. Our no-code platform allows you to generate predictive and prescriptive analytics without needing to write a single line of code.

What industries can benefit from using Graphite Note?

Graphite Note is versatile and can be beneficial to a wide range of industries. This includes but is not limited to retail, e-commerce, marketing, sales, finance, healthcare, and manufacturing. Any industry that relies on data to make informed decisions can benefit from our platform.

How can Graphite Note help with my specific business needs?

Graphite Note is a flexible platform that can be tailored to meet your specific business needs. Whether you're looking to improve customer retention, optimize your marketing campaigns, forecast sales, or identify trends, our platform can provide the insights you need to drive growth.

What training resources are available for new users?

We offer a variety of resources to help new users get started with Graphite Note. This includes step-by-step tutorials, webinars, and a comprehensive knowledge base. We're committed to helping you get the most out of our platform and will work with you during onboarding.

How to update a model with new data?

What is a tag?

A tag is a keyword associated with a model or dataset. It is a tool to group your models to easily find them, as you can filter your list by tags.

What does parsing mean?

In programming, parsing refers to the process of analyzing and interpreting the structure of a sequence of characters or symbols according to a specific grammar or syntax. It is used in our application to understand and extract meaningful information from input data.

During parsing, a parser takes the input, which can be a program's source code or any other form of textual data, and breaks it down into a hierarchical structure that conforms to a predefined set of rules or grammar. This hierarchical structure is typically represented using a data structure such as an abstract syntax tree (AST) or a parse tree.

How does the free trial work?

Users can start their free trial of the SPROUT plan immediately. The trial is valid for the next 14 days. After 14 days, if you want to continue our service, you must subscribe to a plan in communication with our sales team.

What's the difference between the starter (SPROUT) and other plans?

The starter plan is primarily designed for individual users that want to upload CSV files and create machine learning models.

The starter plan has the same core functionality as higher plans but with the following limitations:

• Only one user in the workspace.

• Only CSV connector

• Max of 3 models can be created

• Number of total data source rows limited to 1 milion.

What is API access?

What is a Dedicated Data Scientist?

With Graphite Note, you have the option to add a Dedicated Data Scientist to your team. This is an expert in machine learning and data science who can assist you and your team with any questions or concerns you may have. They can also provide hands-on support with tasks such as data cleaning and improving the performance of your models.

Can I extend my trial?

We can extend your trial beyond the one-week default period in certain circumstances. Don't hesitate to get in touch with us before the end of your trial if you'd like to discuss this further.

Special Discounts for Academia and Startups?

Our SaaS platform provides no-code predictive analytics, making data analysis accessible to everyone. As a company, we are dedicated to supporting the academic and startup communities and offer generous discounts to these groups.

If you want to hear more about our offerings, don’t hesitate to reach out to us!

Reputable Provider

Graphite Note runs on platforms belonging to reputable leading service providers and vendors that uphold the highest security standards, specifically: Amazon Web Service (AWS).

I want to know more about Data Science and Machine Learning lexical terms

How is Graphite Note different then Gen AI?

In summary, Graphite Note is a specialized tool for predictive analytics and data-driven decision-making, whereas Generative AI focuses on creating new content based on prompts. Both have unique strengths, but they serve to vastly different business and creative needs.

Graphite Note specializes in machine learning tasks like regression, classification, and actionable insights based on structured datasets. It’s tailored for business scenarios such as sales forecasting, churn prediction, or lead scoring, Generative AI excels at generating unstructured outputs such as creative writing, dialogue generation, or designing visuals based on prompts.

Introduction

In predictive modeling, key drivers (or influencers) are pivotal in discerning which features within a dataset most significantly impact the target variable. These influencers provide insights into the relative importance of each variable, enabling data scientists and analysts to understand and predict outcomes more accurately.

By highlighting the strongest predictors, key influencers inform the prioritization of features for model optimization, ensuring that models are precise and interpretable in real-world scenarios. This foundational understanding is crucial for refining models and aligning them closely with the underlying patterns and trends present in the data.

Reading Key Drivers

When examining the visualization of key influencers in Graphite Note Models, you'll find features arrayed according to their influence on the target variable, organized from most to least important on the left.

This ranking allows for a quick assessment of which factors are pivotal in the model's predictions.

By observing the length and direction of the bars associated with each feature, one can gauge the strength of influence they have on the target outcome.

The image shows a data visualization explaining how different amounts of interaction with website pages (measured in page visits) influence whether someone will take a specific action, labeled "Applied," with "YES" being the action taken.

1. For a high number of page visits, between 29.33 and 35, the likelihood of taking the action increases significantly—by more than double (2.26 times more likely).

2. For a moderate number of page visits, between 12.33 and 18, the action is still more likely but less so than the higher range—1.65 times more likely.

3. At a lower number of page visits, between 6.67 and 12.33, the action becomes less likely than the baseline by a factor of 1.37.

4. For very few page visits, less than 6.67, the likelihood of action drops drastically to less than half (2.36 times less likely).

The percentages and observations indicate how many cases fall within each range and how many of those cases resulted in the action "Applied" being taken. The visualization communicates that more engagement with the website (as measured by page visits) generally increases the likelihood of the desired action occurring.

Statistical Methodology Used

Graphite Note uses advanced statistical functions designed to calculate the influence of features on a target variable.

It employs a method of grouping the data by the feature and target columns and then counting occurrences. The calculations performed within this function aim to determine the proportion of each feature's categories contributing to a specific target value. The influence is quantified by comparing the observed proportion of the target value within each feature category against a weighted average, yielding an 'index value' that indicates the relative influence of each category on the target outcome. The function is robust, allowing for different data types in the target column, and ensures that only relevant categories with sufficient data are included in the final analysis.

Graphite Note here a quantitative analysis where numeric features (like 'Website Pages') are divided into bins or ranges.

The function then calculates the change in the likelihood of the target outcome (e.g., 'Applied' being 'YES') when the feature values fall within those bins. This calculation is done by comparing the base likelihood of the target outcome with the likelihood when the feature is within a specific bin, hence the multipliers like "increases by 2.26x" for certain ranges.

The analysis would remove any non-relevant categories (based on minimum percentage and row thresholds) and sort the results to clearly show which ranges of the feature increase or decrease the likelihood of the target outcome.

Data to Insights Lifecycle

• Dataset: Begin with a dataset containing historical data.

• Feature Selection: Identify the most important variables (features) for the model.

• Best Algorithm Search: Test different algorithms to find the best fit for your data.

• Model Generation: Create a predictive model based on selected features and the best algorithm.

• Model Tuning: Fine-tune the model’s parameters to improve accuracy.

• Model Deployment: Deploy the final model for real-world usage.

• Explore Key Drivers: Analyze the key factors influencing the model’s predictions.

• Explore What-If Scenarios: Test different hypothetical situations to see their impact.

• Predict Future Outcomes: Use the model to forecast future trends or outcomes.

Purpose

The Confusion Matrix is a powerful diagnostic tool in classification tasks within predictive analytics. It presents a clear and concise layout for evaluating the performance of a classification model by showing the actual versus predicted values in a tabular format. The matrix allows users, regardless of their coding expertise, to assess the accuracy and effectiveness of a predictive model, providing insights into not only the number of correct and incorrect predictions but also the type of errors made.

Components of the Confusion Matrix

A confusion matrix for a binary classification problem consists of four components:

• True Positives (TP): The number of instances that were predicted as positive and are actually positive.

• False Positives (FP): The number of instances that were predicted as positive but are actually negative.

• True Negatives (TN): The number of instances that were predicted as negative and are actually negative.

• False Negatives (FN): The number of instances that were predicted as negative but are actually positive.

Applications

In the context of Graphite Note, a no-code predictive analytics platform, the confusion matrix serves several key purposes:

1. Performance Measurement: It quantifies the performance of a classification model, offering a visual representation of the model's ability to correctly or incorrectly predict categories.

2. Error Analysis: By breaking down the types of errors (FP and FN), the matrix aids in understanding specific areas where the model may require improvement.

3. Decision Support: The confusion matrix supports decision-making by highlighting the balance between sensitivity (or recall) and precision, which can be crucial for business outcomes.

4. Model Tuning: Users can leverage the insights from the confusion matrix to adjust model parameters and thresholds to optimize for certain predictive behaviors.

5. Communication Tool: It acts as a straightforward communication tool for stakeholders to grasp the results of a classification model without delving into complex statistical jargon.

Interpretation

In the example confusion matrix (Model Performance -> Accuracy Overview)

• There are 799 instances where the model correctly predicted the positive class (TP).

• There are 15622 instances where the model incorrectly predicted the positive class (FP).

• There are 348 instances where the model failed to identify the positive class (FN).

• There are 18159 instances where the model correctly identified the negative class (TN).

The high number of FP and FN relative to TP suggests a potential imbalance or a need for model refinement to improve predictive accuracy.

Conclusion

The classification confusion matrix is an integral part of the model evaluation in Graphite Note, enabling users to make informed decisions about the deployment and iteration of their predictive models.

1. If you want to use Graphite Note demo datasets click "Import DEMO Dataset".

2. Select the dataset you want to use to create a machine learning model. In this case we will select "Ads" dataset to create a Regression Analysis on marketing ads data.

3. Once selected, the demo dataset will load directly to your account. The dataset view will automatically open.

4. Adjust your dataset options on the Settings tab. Click Columns tab to view list of available columns with their corresponding data types. Explore dataset details on Summary tab.

5. To create a new model in the Graphite Note main menu click on "Models".

6. You will get list of available models. Click on "New Model" to create a new one.

7. Select model type from our templates. In our case, we will select "Regression" by double clicking on its name.

8. Select dataset you want to use to produce a model. We will use "Demo-Ads.csv."

9. Name your new model. We will call it "Regression on Demo-Ads".

10. Write the description of the model and select tag. If you want you can also create a new tag from pop-up "Tags" window that will appear on the screen.

11. Click "Create" to create your demo model environment.

12. To set up a "Regression Model", first you will need to define the "Target Feature". That is a numeric column from your dataset that you'd like to make predictions about. In the case of Regression on Ads dataset, target feature is "Clicks" column.

13. Click "Next" to get the list of model features that will be included in the scenario. Model relies upon each column (feature) to make accurate predictions. When training the model, we will calculate which of the features are most important and behave as Key Drivers.

14. To start training the model, click "Run scenario". This will take a sample of 80% of your data and train several machine learning models.

15. Wait for few moments and Voilà! Your Regression model is trained. Click on "Performance" tab to get model insights and view Key Drivers.

16. Explore the Regression Model by clicking on Impact Analysis, Model Fit and Training Results to get more insights on how the model is trained and set up.

17. If you want to take turn model into action click on "Predict" tab in the main model menu.

18. You can produce your own "What-If" analysis based on existing training results. You can also import a fresh CSV dataset into the data model, to make predictions on target column. In our case that is "Clicks". Keep in mind, the dataset you are uploading needs to contain same feature columns as your model.

19. Use your model often to predict future behaviour, and to learn which key drivers are impacting the outcomes. The more you use and retrain your model, the smarter it becomes!

1. If you want to use Graphite Note demo datasets click "Import DEMO Dataset".

2. Select the dataset you want to use to create machine learning model. In this case we will select a Diamonds dataset to create Multi Class Analysis on diamond characteristics data.

3. Once selected, the demo dataset will load directly to your account. Dataset view will automatically open.

4. Adjust your dataset options on Settings tab. Click Columns tab to view list of available columns with their corresponding data types. Explore dataset details on Summary tab.

5. To create new model in the Graphite Note main menu click on "Models"

6. You will get list of available models. Click on "New Model" to create a new one.

7. Select a model type from our templates. In our case we will select "Multi-Class Classification" by double clicking on its name.

8. Select the dataset you want to use to produce model. We will use "Demo-Diamonds.csv"

9. Name your new model. We will call it "Multi-Class Classification on Demo-Diamonds"

10. Write description of the model and select tag. If you want you can also create new tag from pop-up "Tags" window that will appear on the screen.

11. Click "Create" to create your demo model environment.

12. To set up a Multi Class model first you need to define "Target Feature". That is text column from your dataset you'd like to make predictions about. In case of Multi Class on Diamonds dataset target feature is "Cut" column.

13. Click "Next" to get the list of model features that will be included under the scenario. Model relies on each column (feature) to make accurate predictions. When training the model, we will calculate which of the features are most important and behave as Key Drivers.

14. To start training model click "Run Scenario". This will take a sample of 80% of your data and train several machine learning models.

15. Wait for few moments and Voilà! Your Regression model is trained. Click on "Performance" tab to get model insights and view Key Drivers.

16. Explore Multi Class model by clicking on Impact Analysis, Model Fit, Accuracy Overview or Training Results to get more insights on how model is trained and set up.

17. If you want to take your model into action click on "Predict" tab in the main model menu.

18. You can produce your own What-If analysis based on existing training results. You can also import a fresh CSV dataset to make predictions on target column. In our case that is "Cut". Keep in mind, the dataset you are uploading needs to contain the same feature columns as your model.

19. Use your model often to predict future behaviour and to learn which key drivers are impacting the outcomes. The more you use and retrain your model, the smarter it becomes!

In machine learning, supervised and unsupervised learning are two main types of approaches:

1. Supervised Learning

In supervised learning, the model is trained on a labeled dataset. This means we provide both input data and the corresponding output labels to the model. The goal is for the model to learn the relationship between inputs and outputs so it can predict new, unseen data. Common examples include classification (e.g., email spam detection) and regression (e.g., predicting house prices).

For example, if you have an image dataset labeled with “cat” or “dog,” the model learns to classify new images as either a cat or dog based on this training.

Supervised learning on Diamonds dataset example

In this example, we have a dataset containing information about diamonds. The supervised machine learning approach focuses on predicting a specific target column based on other features in the dataset.

• Target is a Number (Regression): If the target column is numerical (e.g., “Price”), the goal is to predict the diamond’s price based on features like cut, color, and clarity. This is called regression.

• Target is Text (Classification): If the target is categorical (e.g., “Cut” with values like Ideal, Very Good), the goal is to classify diamonds into categories based on their characteristics. This is known as classification.

2. Unsupervised Learning

In unsupervised learning, the model is given only input data without any labeled outputs. The goal is to find patterns or groupings within the data. A common task here is clustering (e.g., grouping customers by purchasing behavior) and dimensionality reduction (e.g., simplifying data visualization).

For example, if you have images without labels, the model could group similar images together (like cats in one group and dogs in another) based on visual similarities.

Unsupervised learning on Diamonds dataset example

In unsupervised learning, there is no target column or labeled output provided. Instead, the model analyzes patterns within the data to group or cluster similar items together.

In this diamond dataset example:

• We don’t specify a target column (like price or cut); instead, the goal is to find natural groupings of data points based on their features.

• Here, clustering is used to identify groups of diamonds with similar Carat Weight and Price characteristics.

• The scatter plot on the right shows how the diamonds are grouped into different clusters (e.g., cluster0, cluster1, etc.), revealing patterns in the data without needing predefined labels.

This approach is useful when you want the model to identify hidden structures or patterns within the data. Unsupervised learning is often used for customer segmentation, anomaly detection, and recommendation systems.

1. If you want to use Graphite Note demo datasets click "Import DEMO Dataset".

2. Select the dataset you want to use to create the machine learning model. In this case, we will select CO2 Car Emissions dataset to create Regression Analysis on car emissions data.

3. Once selected, the demo dataset will load directly to your account. The dataset view will automatically open.

4. Adjust your dataset options on Settings tab. Click the Columns tab to view list of available columns with their corresponding data types. Explore dataset details on Summary tab.

5. To create a new model in the Graphite Note main menu click on "Models"

6. You will get list of available models. Click on "New Model" to create new one.

7. Select model type from our templates. In our case we will select "Regression" by double clicking on its name.

8. Select the dataset you want to use to produce the model. We will use "Demo-CO2-Car-Emissions-Canada.csv"

9. Name your new model. We will call it "Regression on Demo-CO2-Car-Emissions"

10. Write the model description and select tag. If you want you can also create new tag from pop-up "Tags" window that will appear on the screen.

11. Click "Create" to create your demo model environment.

12. To set up the Regression model first, you need to define "Target Feature". That is numeric column from your dataset that you'd like to make predictions about. In case of Regression on car emissions dataset target feature is "CO2 Emissions(g/km)" column.

13. Click "Next" to get the list of model features that will be included into scenario. Model relies upon each column (feature) to make accurate predictions. When training the model, we will calculate which of the features are most important and behave as Key Drivers.

14. To start training the model, click "Run scenario". This will take a sample of 80% of your data and train several machine learning models.

15. Wait for few moments and Voilà! Your Regression model is trained. Click on the "Performance" tab to get model insights and view the Key Drivers.

16. Explore the Regression model by clicking on Impact Analysis and Training Results, to get more insights on how model is trained.

17. If you want to take your model into action, click on the "Predict" tab in the main model menu.

18. You can produce your own What-If analysis based on existing training results. You can also import a fresh CSV dataset that the model will use to make predictions on the target column. In our case that is "CO2 Emissions (g/km)". Keep in mind, the dataset you are uploading needs to contain same feature columns as your model.

19. Use your model often to predict future behaviour and to learn which key drivers are impacting the outcomes. The more you use and retrain your model, the smarter it becomes!

We have prepared 12 distinct demo datasets that you can upload and use in your Graphite Note account. These demo datasets include dummy data from a variety of business scenarios and serve as an excellent starting point for building and running your initial Graphite Note machine learning models. Instructions on how to proceed with demo dataset will be provided in the following pages.

The "What Dataset Do I Need?" section of Graphite Note is a comprehensive resource designed to guide users through the intricacies of dataset selection and preparation for various machine learning models. This section is crucial for users, especially those without extensive AI expertise, as it provides clear, step-by-step instructions and examples on how to curate and structure data for different predictive analytics scenarios.

Key Features of the Section

1. Model-Specific Guidance: Each page within this section is tailored to a specific predictive model, such as cross-selling prediction, churn prediction, or customer segmentation. It outlines the type of data required, the format, and how to interpret and use the data effectively.

2. Sample Datasets and Templates: To make the process more user-friendly, the section includes sample datasets and templates. These examples showcase the necessary columns and data types, along with a brief explanation of each, helping users to model their datasets accurately.

3. Target Column Identification: A crucial aspect of preparing a dataset for machine learning is identifying the target column. This section provides clear guidance on selecting the appropriate target for different types of analyses, whether it's for classification, regression, or clustering.

4. Data Cleaning and Preparation Tips: Recognizing that data rarely comes in a ready-to-use format, this section offers valuable tips on cleaning and preparing data, ensuring that users start their predictive analytics journey on the right foot.

5. Real-World Applications and Use Cases: To bridge the gap between theory and practice, the section includes examples of real-world applications and use cases. This approach helps users understand how their data preparation efforts translate into actionable insights in various business contexts.

1. If you want to use Graphite Note demo datasets click "Import DEMO Dataset".

2. Select the dataset you want to use to create machine learning model. In this case we will select eCommerce Orders dataset to create RFM Customer Segmentation (Recency, Frequency, Monetary Value) analysis on ecommerce orders data.

3. Once selected, the demo dataset will load directly to your account. Dataset view will automatically open.

4. Adjust your dataset options on Settings tab. Click Columns tab to view list of available columns with their corresponding data types. Explore the dataset details on Summary tab.

5. To create new model in the Graphite Note main menu click on "Models"

6. You will get list of available models. Click on "New Model" to create new one.

7. Select a model type from our templates. In our case we will select "RFM Customer Segmentation" by double clicking on its name.

8. Select dataset you want to use to produce model. We will use "Demo-eCommerce-Orders.csv".

9. Name your new model. We will call it "RFM customer segmentation on Demo-eCommerce-Orders".

10. Write description of the model and select tag. If you want you can also create new tag from pop-up "Tags" window that will appear on the screen.

11. Click "Create" to create your demo model environment.

12. Click this text field.

13. To set up RFM model, you first need to identify and define few parameters. These are: "Time /Date Column", "Customer ID", "Customer Name" (optional) and "Monetary" (amount spent). In our case we will select "created_at" as date, "user_id" as customer and "total" as monetary parameter.

14. To start training model click "Run scenario".

15. Wait for a few moments and Voilà! Your RFM Customer Segmentation model is trained. Click on the "Results" tab to get model insights.

16. You can navigate over different tabs to get deep insights into RFM analysis from different perspectives: Recency, Frequency, Monetary.

17. Tab "RFM Scores" shows detailed explanation on different scores along with RFM segments and descriptions.

18. Tab "RFM Analysis" gives you more details on different segments

19. Tab "RFM Matrix" will show you number of customers belonging to different RFM segment. You can export matrix data to use Customer IDs for different business actions (e.g. exporting list of about to churn customers).

1. If you want to use Graphite Note demo datasets click "Import DEMO Dataset".

2. Select a dataset you want to use to create machine learning model. In this case we will select Housing-Prices dataset to create a "Regression Analysis" on house price historical data.

3. Once selected, the demo dataset will load directly to your account. The Dataset view will automatically open.

4. Adjust your dataset options on the Settings tab. Click Columns tab to view list of available columns with their corresponding data types. Explore the dataset details on the Summary tab.

5. To create a new model in the Graphite Note main menu click on "Models"

6. You will get list of available models. Click on a "New Model" to create new one.

7. Select the model type from our templates. In our case, we will select "Regression" by double clicking on its name.

8. Select the dataset you want to use to produce the model. We will use "Demo-Housing-Prices.csv"

9. Name your new model. We will call it "Regression on Demo-Housing-Prices"

10. Write the description of the model and select a tag. If you want to, you can also create new a tag from pop-up "Tags" window that will appear on the screen.

11. Click "Create" to create your demo model environment.

12. To set up a Regression Model, firstly, you need to define the "Target Feature". That is a numeric column from your dataset that you'd like to make predictions about. In the case of Regression on Demo Housing Prices, the dataset target feature is "Price" column.

13. Click "Next" to get the list of model features that will be included into scenario. Model relies upon each column (feature) to make accurate predictions. When training model we will calculate which of the features are most important and behave as Key Drivers.

14. To start training the model, click "Run scenario". This will take a sample of 80% of your data and train several machine learning models.

15. Wait for few moments and Voilà! Your Regression model is trained. Click on the "Performance" tab to get model insights and view the Key Drivers.

16. Explore the Regression Model by clicking on Impact Analysis, Model Fit and Training Results to get more insights on how model is trained and set up.

17. If you want to take your model into action. Click on "Predict" tab in the main model menu.

18. You can produce your own What-If analysis based on existing training results. You can also import fresh CSV dataset with data model will use to make predictions on the target column. In our case, that is "Price". Keep in mind, the dataset you are uploading needs to contain the same feature columns as your model.

19. Use your model often to predict future behaviour, and to learn which key drivers are impacting the outcomes. The more you use and retrain your model, the smarter it becomes!

Get an overview of Lead Scoring demo dataset and how it can be used to create your new Graphite Note model in this video:

Or follow instructions below to get step by step guidance on how to use Lead Scoring demo dataset:

1. If you want to use Graphite Note demo datasets click "Import DEMO Dataset".

2. Select the dataset you want to use to create the machine learning model. In this case, we will select "Lead Scoring dataset" to create binary classification analysis on potential customer interactions data.

3. Once selected, the demo dataset will load directly to your account. The Dataset view will automatically open.

4. Adjust your dataset options on the Settings tab. Click Columns tab to view the list of available columns with their corresponding data types. Then explore the dataset details on Summary tab.

5. Click "Models"

6. You will get list of available models. Click on "New Model" to create a new one.

7. Select the model type from our templates. In our case, we will select "Binary Classification" by double clicking on its name.

8. Select dataset you want to use to produce the model. We will use "Demo-Lead-Scoring.csv."

9. Name your new model. We will call it "Binary Classification on Demo-Lead-Scoring".

10. Write the description of the model and select a tag. If you want to, you can also create a new tag from pop-up "Tags" window that will appear on the screen.

11. Click "Create" to create your demo model environment.

12. To set up a Binary Classification model, firstly, you need to define the "Target Feature". That is a binary column from your dataset that you'd like to make predictions about. In the case of Binary Classification on a Lead Scoring dataset, the target feature will be the "Converted" column.

13. Click "Next" to get the list of model features that will be included in the model scenario. The model relies upon each column (feature) to make accurate predictions. When training the model, it will calculate which of the features are most important and behave as Key Drivers.

14. To start training model click "Run Scenario". This will take a sample of 80% of your data and train several machine learning models.

15. Wait for few moments and Voilà! Your Binary Classification model is trained. Click on the "Performance" tab to get model insights and to view the Key Drivers.

16. Explore the "Binary Classification" model by clicking on the Impact Analysis and Training Results to get more insights on how the model is trained.

17. If you want to turn your model into action, click on "Predict" tab in the main model menu.

18. You can produce your own What-If analysis based on existing training results. You can also import a fresh CSV dataset to make predictions on the target column. In our case that is "Converted". Keep in mind, dataset you are uploading needs to contain same feature columns as your model.

19. Use your model often to predict future behaviour, and to learn which key drivers are impacting outcomes. The more you use and retrain your model, the smarter it becomes!

1. If you want to use Graphite Note demo datasets click "Import DEMO Dataset".

2. Select a dataset you want to use to create your advanced analytics model. In this case, we will select Monthly Car Sales dataset to create a "Timeseries Forecast" analysis on car sales data.

3. Once selected, the demo dataset will load directly to your account. Dataset view will automatically open.

4. Adjust your dataset options on the Settings tab. Click the Columns tab to view the list of available columns with their corresponding data types. Explore the dataset details on the Summary tab.

5. To create a new model in the Graphite Note main menu, click on "Models".

6. You will get list of available models. Click on "New Model" to create new one.

7. Select the model type from our templates. In our case, we will select "Timeseries Forecast" by double clicking on its name.

8. Select the dataset you want to use to produce the model. We will use "Demo-Monthly-Car-Sales.csv".

9. Name your new model. We will call it "Timeseries forecast on Demo-Monthly-Car-Sales".

10. Write description of the model and select a tag. If you want to, you can also create new tag from pop-up "Tags" window, that will appear on the screen.

11. Click "Create" to create your demo model environment.

12. To set up Timeseries forecast analysis first you need to define the "Target Column". That is a numeric column from your dataset that you'd like to forecast. In the case of Timeseries on monthly car sales dataset target column is "Sales"

13. If dataset includes multiple time series sequences, you can select field that will be used to uniquely identify each sequence. In the case of our demo dataset, we will not apply Sequence Identifier field since we have only "Sales" target column.

14.

15. Click "Next" to open "Time/Date Column" selection. Choose "Month" as date column.

16. From additional options below, choose "Monthly" as time interval and define "Forecast Horizon". We will set up forecast horizon to 6 months in the future.

17. Click "Next" to activate "Seasonality" options step. Here, you can define seasonality specifics of your forecast. If time interval is set to daily on the next step you will also have "Advanced options" available.

18. Click "Run Scenario" to train your timeseries forecast.

19. Wait for a few moments and Voilà! Your Timeseries forecast is trained. Click on the "Performance" tab to get insights and view the graph with original(historical) and predicted model data.

20. Explore more details on "trend", "Seasonality" and "Details" tabs.

21. If you want to turn your model into action click on "Predict" tab in the main model menu.

22. You can produce your own Forecast analysis based on the existing training results by selecting Start and End date from drop down calendar and clicking on "Predict" button.

23. Use your model often to predict future sales results. The more you use and retrain your model, the smarter it becomes!

1. If you want to use Graphite Note demo datasets click "Import DEMO Dataset".

2. Select dataset you want to use to create your machine learning model. In this case we will select Upsell dataset to create binary classification analysis on additional purchases by customer data.

1. After selection, the demo dataset will automatically load into your account and the dataset view will open immediately.

4. Adjust your dataset options on the Settings tab. Click Columns tab to view the list of available columns with their corresponding data types. Explore dataset details on Summary tab.

5. To create a new model in the Graphite Note main menu, click on "Models"

6. You will get a list of available models. Click on "New Model" to create a new one.

7. Select the model type from our templates. In our case we will select "Binary Classification" by double clicking on its name.

8. Select the dataset you want to use to produce model. We will use "Demo-Upsell.csv".

9. Name your new model. We will call it "Binary Classification on Demo-Upsell".

10. Write a description of the model and select a tag. If you want you can also create new tag from pop-up "Tags" window that will appear on the screen.

11. Click "Create" to create your demo model environment.

12. To set up a Binary Classification model, firstly, you need to define "Target Feature". That is binary column from your dataset that you'd like to make predictions about. In case of Binary Classification on Upsell dataset target feature will be "Applied" column.

13. Click "Next" to get the list of model features that will be included in the model scenario. Your model relies upon each column (feature) to make accurate predictions. When training the model we will calculate which of the features are most important and behave as Key Drivers.

14. To start training your model click "Run scenario". This will take a sample of 80% of your data and train several machine learning models.

15. Wait for few moments and Voilà! Your Binary Classification model is trained. Click on "Performance" tab to get model insights and view the Key Drivers.

16. Explore Binary Classification model by clicking on Impact Analysis and Training Results to get more insights on how model is trained.

17. If you want to take your model into action click on "Predict" tab in the main model menu.

18. You can produce your own What-If analysis based on existing training results. You can also import a fresh CSV dataset to make predictions on the target column. In our case that is "Applied". Keep in mind, the dataset you are uploading needs to contain the same feature columns as your model.

19. Use your model often to predict future behaviour and to learn which key drivers are impacting the outcomes. The more you use and retrain your model, the smarter it becomes!

Get an overview of Customer Churn demo dataset and how it can be used to create your new Graphite Note model in this video:

Or follow instructions below to get step by step guidance on how to use Customer Churn demo dataset:

1.If you want to use Graphite Note demo datasets click "Import DEMO Dataset"

2. Select the dataset you want to use to create a machine learning model. In this case we will select Churn dataset to create binary classification analysis on customer engagement data .

3. Once selected, demo dataset will load into your account. Dataset view will automatically open.

4. Adjust your dataset options on Settings tab. Click Columns tab to view list of available columns with their corresponding data types. Explore dataset details on Summary tab.

5. To create new model in the Graphite Note main menu click on "Models"

6. You will get list of available models. Click on "New Model" to create new one.

7. Select model type from our templates. In our case we will select "Binary Classification" by double clicking on its name.

8. Select dataset you want to use to produce model. We will use "Demo-Churn.csv."

9. Name your new model. We will call it "Binary Classification on Demo-Churn".

10. Write description of the model and select tag. If you want to, you can also create a new tag from pop-up "Tags" window that will appear on the screen.

11. Click "Create" to create your demo model environment.

12. To set up Binary Classification model first you need to define "Target Feature". That is binary column from your dataset that you'd like to make predictions about. In case of Binary Classification on Churn dataset, the target feature will be "Churn" column.

13. Click "Next" to get the list of model features that will be included in scenario. Model relies upon each column (feature) to make accurate predictions. When training model we will calculate which of the features are most important and behave as Key Drivers.

14. To start training model click "Run scenario". This will take a sample of 80% of your data and train several machine learning models.

15. Wait for a few moments and Voilà! Your Binary Classification model is trained. Click on "Performance" tab to get model insights and view Key Drivers.

16. Explore Binary Classification model by clicking on Impact Analysis and Training Results to get more insights on how model is trained.

17. If you want to turn your model into action click on "Predict" tab in the main model menu.

18. You can produce your own "What-If analysis" based on existing training results. You can also import a fresh CSV dataset with data model will use to make predictions on a target column. In our case that is "Churn". Keep in mind, the dataset you are uploading needs to contain same feature columns as your model.

19. Use your model often to predict future behaviour and to learn which key drivers are impacting the outcomes. The more you use and retrain your model, the smarter it becomes!

1. If you want to use Graphite Note demo datasets click "Import DEMO Dataset".

2. Select the dataset you want to use to create your machine learning model. In this case, we will select MMM dataset to create Regression Analysis on marketing mix and sales data.

3. Once selected, the demo dataset will load directly to your account. The dataset view will automatically open.

4. Adjust your dataset options on the Settings Tab. Click the Columns tab to view list of available columns with their corresponding data types. Explore dataset details on the Summary tab.

5. To create a new model in the Graphite Note main menu click on "Models".

6. You will get list of available models. Click on "New Model" to create a new one.

7. Select your model type from our templates. In our case we will select "Regression" by double clicking on its name.

8. Select dataset you want to use to produce the model. We will use "Demo-MMM.csv"

9. Name your new model. We will call it "Regression on Demo-MMM".

10. Write description of the model and select tag. If you want you can also create new tag from pop-up "Tags" window that will appear on the screen.

11. Click "Create" to create your demo model environment.

12. To set up your "Regression Model", firstly, you need to define "Target Feature". That is numeric column from your dataset that you'd like to make predictions about. In the case of Regression on Marketing Mix and Sales Dataset, the target feature is "Sales" column.

13. Click "Next" to get the list of model features that will be included in scenario. The model relies upon each column (feature) to make accurate predictions. When training the model, we will calculate which of the features are most important and behave as Key Drivers.

14. To start training the model, click "Run Scenario". This will take a sample of 80% of your data and train several machine learning models.

15. Wait for few moments and Voilà! Your Regression model is trained. Click on "Performance" tab to get model insights and view the Key Drivers.

16. Explore Regression model by clicking on Impact Analysis, Model Fit and Training Results to get more insights on how the model is trained and set up.

17. If you want to take your model into action, click on "Predict" tab in the main model menu.

18. You can produce your own What-If analysis based on existing training results. You can also import fresh CSV dataset with data model will use to make predictions on target column. In our case, that is "Sales". Keep in mind, dataset you are uploading needs to contain same feature columns as your model.

19. Use your model often to predict future behaviour and to learn which key drivers are impacting the outcomes. The more you use and retrain your model, the smarter it becomes!

1. If you want to use Graphite Note demo datasets click "Import DEMO Dataset".

2. Select the dataset you want to use to create your machine learning model. In this case, we will select "Mall Customers dataset", to create General Segmentation analysis on customer engagement data.

3. Once selected, the demo dataset will load directly to your account. The dataset view will automatically open.

4. Adjust your dataset options on the Settings tab. Click the Columns tab to view the list of available columns, with their corresponding data types. Explore the dataset details on Summary tab.

5. To create a new model in the Graphite Note main menu, click on "Models"

6. You will get list of available models. Click on the "New Model" to create a new one.

7. Select the model type from our templates. In our case, we will select "General Segmentation" by double clicking on its name.

8. Select the dataset you want to use to produce model. We will use "Demo-Mall-Customers.csv"

9. Name your new model. We will call it "General Segmentation on Demo-Mall-Customers".

10. Write a description of the model and select a tag. If you want to, you can also create new tag from the pop-up "Tags" window that will appear on the screen.

11. Click "Create" to create your demo model environment.

12. To set up your General Segmentation model, firstly, you need to define "Feature" columns. That is numeric column (or columns) from your dataset on which segmentation would be based. In the case of General Segmentation on Mall Customers dataset, the numeric feature will be "Age", Annual Income" and "Spending Score" columns.

13. To start training the model click "Run Scenario". This will train your model based on the uploaded dataset.

14. Wait for few moments and Voilà! Your General Segmentation model is trained. Click on "Results" tab to get model insights and explore segmentation clusters.

15. Navigate over different tabs to get insights from high level "Cluster Summary" to "By Cluster" charts and tables.

16. Use "Cluster Visualisations" to view the scatter plot visualisations of cluster members.

1. If you want to use Graphite Note demo datasets click "Import DEMO Dataset".

2. Select dataset you want to use to create your machine learning model. In this case, we will select Store Item Demand dataset to create Regression analysis on sales across store locations data.

3. Once selected, the demo dataset will load directly to your account. The dataset view will automatically open.

4. Adjust your dataset options on Settings tab. Click Columns tab to view list of available columns with their corresponding data types. Explore dataset details on Summary tab.

5. To create a new model in the Graphite Note main menu, click on "Models".

6. You will get list of available models. Click on "New Model" to create new one.

7. Select model type from our templates. In our case we will select "Regression" by double clicking on its name.

8. Select the dataset you want to use to produce model. We will use "Demo-Store-Item-Demand.csv".

9. Name your new model. We will call it "Regression on Demo-Store-Item-Demand".

10. Write a description of the model and select tag. If you want to, you can also create new tag from pop-up "Tags" window that will appear on the screen.

11. Click "Create" to create your demo model environment.

12. To set up Regression Model, firstly, you will need to define "Target Feature". That is the numeric column from your dataset that you'd like to make predictions about. In case of Regression on Store Item Demand dataset target feature is "Sales" column.

13. Click "Next" to get the list of model features that will be included in model scenario. The model relies on each column (feature) to make accurate predictions. When training a model, we will calculate which of the features are most important and behave as the Key Drivers.

14. To start training your model click "Run scenario". This will take a sample of 80% of your data and train several machine learning models.

15. Wait for few moments and Voilà! Your Regression model is trained. Click on "Performance" tab to get model insights and view Key Drivers.

16. Explore Regression model by clicking on Impact Analysis, Model Fit and Training Results to get more insights on how model is trained and set up.

17. If you want to take your model into action click on "Predict" tab in the main model menu.

18. You can produce your own What-If analysis based on existing training results. You can also import a fresh CSV dataset to make predictions on target column. In our case that is "Sales". Keep in mind, dataset you are uploading needs to contain same feature columns as your model.

19. Use your model often to predict future behaviour and to learn which key drivers are impacting the outcomes. The more you use and retrain your model, the smarter it becomes!

Predict Revenue is a critical task for businesses aiming to forecast future revenue streams accurately. This challenge is typically addressed using a time series forecasting model, which analyzes historical revenue data to predict future trends and patterns.

Dataset Essentials for Predict Revenue

A suitable dataset for Predict Revenue using time series forecasting should include:

• Date/Time: The timestamp of revenue data, usually in daily, weekly, or monthly intervals.

• Revenue: The total revenue recorded in each time period.

• Seasonal Factors: Data on seasonal variations or events that might affect revenue.

• Economic Indicators: Relevant economic factors that could influence revenue trends.

• Marketing Spend: Information on marketing and advertising expenditures, if applicable.

An example dataset for Predict Revenue with time series forecasting might look like this:

Target Column: The Total Revenue column is the primary focus, as the model aims to forecast future values in this series.

Steps to Success with Graphite Note

1. Data Collection: Compile historical revenue data along with any relevant external factors.

2. Time Series Analysis: Utilize Graphite Note to analyze the time series data and identify patterns.

3. Model Training: Train a time series forecasting model using the platform.

4. Model Evaluation: Continuously evaluate and adjust the model based on new data and changing market conditions.

Benefits of Predict Revenue with Time Series Forecasting

• Accurate Financial Planning: Enables more precise budgeting and financial planning.

• Strategic Decision Making: Informs strategic decisions with insights into future revenue trends.

• Adaptability to Market Changes: Helps businesses adapt strategies in response to predicted market changes.

• User-Friendly Analytics: Graphite Note's no-code approach makes sophisticated time series forecasting accessible to users without specialized statistical knowledge.

In summary, Predict Revenue with time series forecasting is an essential tool for businesses to anticipate future revenue trends. Graphite Note simplifies this complex task, allowing businesses to leverage their historical data for insightful and actionable revenue predictions.

Product Demand Forecast is a crucial process for businesses to predict future demand for their products. This task typically involves time series forecasting models, which analyze historical sales data to forecast future demand patterns.

Dataset Essentials for Product Demand Forecast

An effective dataset for Product Demand Forecast using time series forecasting should include:

• Date/Time: The timestamp for each data point, typically daily, weekly, or monthly.

• Product Sales: The number of units sold or the sales volume of each product.

• Product Features: Characteristics of the product, such as category, price, or any special features.

• Promotional Activities: Data on any marketing or promotional activities that might affect sales.

• External Factors: Information on external factors like market trends, economic conditions, or seasonal events.

An example dataset for Product Demand Forecast might look like this:

Target Column: The Sales Volume column is the primary focus, as the model aims to forecast future sales volumes for each product.

Steps to Success with Graphite Note

1. Data Collection: Gather detailed sales data along with product features and external factors.

2. Time Series Analysis: Use Graphite Note to analyze the sales data over time, identifying trends and patterns.

3. Model Training: Train a time series forecasting model on the platform.

4. Model Evaluation: Regularly evaluate the model's performance and adjust it based on new data and market changes.

Benefits of Product Demand Forecast

• Inventory Management: Helps in planning inventory levels to meet future demand, avoiding stockouts or overstock situations.

• Strategic Marketing: Informs marketing strategies by predicting when demand for certain products will increase.

• Resource Allocation: Assists in allocating resources efficiently based on predicted product demand.

• Accessible Forecasting: Graphite Note's no-code platform makes advanced forecasting techniques accessible to a wider range of users.

In summary, Product Demand Forecast is vital for businesses to anticipate market demand and plan accordingly. With Graphite Note, this complex analytical task becomes manageable, enabling businesses to leverage their data for effective demand planning and strategic decision-making.

RFM Customer Segmentation: An Overview RFM (Recency, Frequency, Monetary) customer segmentation is a method businesses use to categorize customers based on their purchasing behavior. This approach helps personalize marketing strategies, improve customer engagement, and increase sales.

The segmentation is based on three criteria:

• Recency: How recently a customer made a purchase.

• Frequency: How often they make purchases.

• Monetary Value: How much money they spend.

Essential Dataset Components for RFM Segmentation A robust dataset for effective RFM segmentation includes the following key elements:

1. Date (Recency): The date of each customer's last transaction, essential for assessing the 'Recency' aspect of RFM.

2. Customer ID: A unique identifier for each customer, crucial for tracking individual purchasing behaviors.

3. Monetary Spent (Monetary Value): The total amount spent by the customer in each transaction, to evaluate the 'Monetary' component of RFM.

Example Dataset for RFM Customer Segmentation

Steps to Success with Graphite Note for RFM Segmentation

1. Data Collection: Gather comprehensive data including customer IDs, transaction dates, and amounts spent.

2. Data Analysis: Utilize Graphite Note to dissect the data, focusing on recency, frequency, and monetary values of customer transactions.

3. Segmentation Modeling: Employ models to segment customers based on RFM criteria, facilitating targeted marketing strategies.

Benefits of RFM Segmentation Using Graphite Note

• Enhanced Marketing Strategies: Tailor marketing campaigns based on customer segments.

• Improved Customer Engagement: Customize interactions based on individual customer behaviors.

• Efficient Resource Allocation: Focus efforts on the most profitable customer segments.

• Strategic Business Decisions: Make informed choices regarding customer relationship management and retention strategies.

In conclusion, RFM Customer Segmentation is a powerful approach for businesses seeking to understand and cater to their customers more effectively. Graphite Note offers a no-code platform that simplifies the analysis of customer data for RFM segmentation, enabling businesses to leverage their data for strategic advantage in customer engagement and retention.

Predictive Lead Scoring is a technique used to rank leads in terms of their likelihood to convert into customers. This approach typically employs a binary classification model, where each lead is classified as 'high potential' or 'low potential' based on various attributes and behaviors.

Dataset Essentials for Predictive Lead Scoring

To effectively implement Predictive Lead Scoring, a dataset with the following elements is essential:

• Lead Demographics: Information such as age, location, and job title.

• Engagement Metrics: Data on how the lead interacts with your business, like website visits, email opens, and download history.

• Lead Source: The origin of the lead, such as organic search, referrals, or marketing campaigns.

• Previous Interactions: History of past interactions, including calls, emails, or meetings.

• Purchase History: If applicable, details of past purchases or subscriptions.

An example dataset for Predictive Lead Scoring might look like this:

Target Column: The Converted column is the target variable. It indicates whether the lead converted to a customer.

Steps to Success with Graphite Note

1. Data Collection: Gather detailed and relevant data on leads.

2. Feature Selection: Choose the most predictive features for lead scoring.

3. Model Training: Utilize Graphite Note to train a binary classification model.

4. Model Evaluation: Test and refine the model for optimal performance.

Benefits of Predictive Lead Scoring

• Efficient Lead Management: Prioritize leads with the highest conversion potential, optimizing sales efforts.

• Personalized Engagement: Tailor interactions based on the lead's predicted preferences and potential.

• Resource Optimization: Allocate marketing and sales resources more effectively.

• Accessible Analytics: Graphite Note's no-code platform makes predictive lead scoring accessible to teams without deep technical expertise.

In summary, Predictive Lead Scoring is a powerful tool for optimizing sales and marketing strategies. With Graphite Note, businesses can leverage advanced analytics to score leads effectively, enhancing their conversion rates and overall efficiency.

Predictive Ads Performance is a process where businesses forecast the effectiveness of their advertising campaigns, particularly focusing on metrics like clicks, conversions, or engagement. This task typically involves regression or classification models, depending on the specific goals of the prediction.

Dataset Essentials for Predictive Ads Performance

A comprehensive dataset for Predictive Ads Performance focusing on predicting clicks should include:

• Date/Time: The timestamp for when the ad was run.

• Ad Characteristics: Details about the ad, such as format, content, placement, and duration.

• Target Audience: Information about the audience targeted by the ad, like demographics, interests, or behaviors.

• Spending: The amount spent on each ad campaign.

• External Factors: Any external factors that might influence ad performance, such as market trends or seasonal events.

• Historical Performance Data: Past performance metrics of similar ads.

An example dataset for Predictive Ads Performance with the target column being clicks might look like this:

Target Column: The Clicks column is the primary focus, as the model aims to forecast the number of clicks each ad will receive.

Steps to Success with Graphite Note

1. Data Collection: Compile detailed data on past ad campaigns, including spending, audience, and performance metrics.

2. Feature Engineering: Identify and create features that are most indicative of ad performance.

3. Model Training: Use Graphite Note, Regression Model, to train a model that can predict the number of clicks based on the ad characteristics and other factors.

4. Model Evaluation: Test the model's accuracy and refine it for better performance.

Benefits of Predictive Ads Performance

• Optimized Ad Spending: Predict which ads are likely to perform best and allocate budget accordingly.

• Targeted Campaigns: Tailor ads to the audience segments most likely to engage.

• Performance Insights: Gain insights into what makes an ad successful and apply these learnings to future campaigns.

• Accessible Analytics: Graphite Note's no-code platform makes predictive analytics accessible, enabling businesses to leverage AI for ad performance prediction without needing deep technical expertise.

In summary, Predictive Ads Performance is a valuable tool for businesses looking to maximize the impact of their advertising efforts. With Graphite Note, this advanced capability becomes accessible, allowing for data-driven decisions in ad campaign management.

Data is an essential component of any data modeling and analysis process. The kind of data you need for modeling depends on the specific problem you are trying to solve. In general, the data should be relevant, accurate, and consistent, and it should cover a significant period. In some cases, you may also need to preprocess or transform the data to make it suitable for modeling.

If you are new to using Graphite Note or are looking for some examples to practice with, there are several popular datasets available that you can explore. Some examples include weather data, financial data, social media data, and sensor data. These datasets are often available in open-source repositories or can be downloaded from public sources, such as government websites, social media platforms, or financial databases.

Graphite Note is a powerful tool that allows you to predict, visualize and analyze data in real-time. With the right dataset, you can use Graphite Note to gain valuable insights and make informed decisions about your business or research. Whether you are analyzing financial data to predict market trends or monitoring sensor data to optimize your production processes, our platform can help you make sense of your data and identify patterns that would be difficult to detect otherwise.

While the kind of data you need may vary depending on your specific needs, there are several popular datasets that you can use to practice and explore the capabilities of Graphite Note. With the right dataset and a solid understanding of data modeling and analysis, you can unlock the full potential of Graphite Note and gain insights that will drive your business or research forward.

We have highlighted a few popular datasets so you can get to know our platform better. After that, it's all up to you - collect your data and start having insights and fun!

An education company named “X Education” sells online courses to industry professionals. Many professionals interested in the courses land on their website and browse for courses on any given day—an excellent dataset for Binary Classification, with a target column "Converted" (YES/NO).

1.1 Usage

Use Graphite Note to gain valuable insights into your sales pipeline by identifying which leads are converting to customers and the factors that contribute to their success. With this information, you can optimize your sales strategy and improve your overall conversion rates.

In addition, our tool can also help you predict which new leads are most likely to convert to customers and provide a probability score for each lead. This can enable you to prioritize your sales efforts and focus on the leads with the highest conversion potential.

By leveraging our tool, you can gain a deeper understanding of your sales funnel and take proactive steps to improve your conversion rates, reduce churn, and increase revenue.

1.2 Model Type

1.3 How to Build This Model

To get started, download the provided dataset and upload it to Graphite Note. Once uploaded, create a new Binary Classification model in Graphite Note with the 'Converted' variable as the Target Variable. This will allow you to predict which leads are most likely to convert to customers.

After training the model, explore the insights that it provides, such as the most important features for predicting conversion and the distribution of conversion probabilities. This can help you to gain a better understanding of the factors that contribute to lead conversion and make informed decisions about your sales strategy.

1.4 What-if?

Finally, you can use the model to run a "what-if" scenario by predicting the conversion probability for new leads based on different scenarios or assumptions. This can help you to forecast the impact of changes in your sales approach or marketing efforts and make data-driven decisions.

By following these steps, you can leverage Graphite Note and the provided dataset to gain valuable insights into your sales pipeline, predict lead conversion, and optimize your sales strategy for better results.

A Telco company customer dataset. Each row represents a customer and each column contains the customer’s attributes. The dataset includes information about:

Customers who left the company – that will be our target column, ("Churn").

Services that each customer has signed up for – phone, multiple lines, internet, online security, online backup, device protection, tech support, and streaming TV and movies.

Customer account information – how long they’ve been a customer, contract, payment method, paperless billing, monthly charges, and total charges.

Demographic info about customers – gender, age range, and if they have partners and dependents.

2.1 Usage

Use Graphite Note to gain valuable insights into your customer base and identify which customers are most likely to churn. By analyzing the factors that contribute to churn, you can optimize your retention strategy and reduce customer churn rates.

In addition, our tool can also help you predict which customers are at high risk of churning, and provide a probability score for each customer. This can enable you to take proactive steps to retain those customers with the highest churn risk, such as offering personalized promotions or improving their overall experience.

By leveraging our tool, you can gain a deeper understanding of your customer base and identify opportunities to reduce churn, increase retention rates, and ultimately drive revenue growth. With our predictive churn model, you can make data-driven decisions that lead to more satisfied customers and a stronger business.

2.2 Model Type

2.3 How to Build This Model

To get started, download the provided dataset and upload it to Graphite Note. Once uploaded, create a new Binary Classification model in Graphite Note with the 'Churn' variable as the Target Variable.

This will allow you to predict which customers are most likely to churn.

After training the model, explore the insights that it provides, such as the most important features for predicting churn and the distribution of churn probabilities. This can help you to gain a better understanding of the factors that contribute to customer churn and make informed decisions about your retention strategy.

2.4 What-if?

Finally, you can use the model to run a "what-if" scenario by predicting the churn probability for different groups of customers based on different scenarios or assumptions. This can help you to forecast the impact of changes in your retention approach or customer experience efforts and make data-driven decisions.

By following these steps, you can leverage Graphite Note and the provided dataset to gain valuable insights into your customer base, predict customer churn, and optimize your retention strategy for better results.

The dataset contains monthly data on car sales from 1960 to 1968. It is great for our time series forecast model with which you can predict sales for the upcoming months.

3.1 Usage

Use Graphite Note to gain valuable insights into your business operations and forecast future trends by analyzing time series data. With our advanced forecasting models, you can make informed decisions about your business and optimize your operations for better results.

Our tool enables you to analyze historical data and identify patterns and trends, such as seasonality or cyclical trends. This can help you to forecast future demand or performance and make data-driven decisions about resource allocation, capacity planning, or inventory management.

3.2 Model Type

3.3 How to Build This Model

To get started, download the provided dataset and upload it to Graphite Note. Once uploaded, create a new Timeseries Forecast model in Graphite Note with

• The 'Sales' variable as aTarget Variable

• Time/Date Column: Month

• Time Interval: Monthly

After training the model, explore the insights that it provides, such as identifying patterns, seasonality, and trends. This can help you to forecast future performance, plan resources effectively, and optimize your operations.

3.4 What-if?

Finally, you can use the model to run a "what-if" scenario by predicting future values.

This can help you to forecast the impact of changes in your business operations, such as changes in demand, capacity planning, or inventory management.

By following these steps, you can leverage Graphite Note to gain valuable insights into your business trends, forecast future performance, and optimize your operations for better results. With our advanced time series forecasting models, you can stay ahead of the competition and take advantage of new opportunities as they arise.

This is a demo CSV with orders for an imaginary eCommerce shop. You can use it for Timeseries forecasting, RFM model, Customer Lifetime Value Model, General Segmentation, or New vs Returning Customers model in Graphite.

A demo Mall Customers dataset from Kaggle. Ideal for General customer segmentation in Graphite.

When you open a dataset, you have five different tabs: , , , , and tabs.

Settings

First, on the Settings tab, you can re-upload the dataset, rename it, and change the description and the tag. You also have the information on the type, the ID, the creation date, and the updated date.

Columns

The Columns tab provides you the original name, the column name, the data type, and the data format of each column, that you can modify.

View Data

On the View Data tab, you have all the data with the number of columns and rows.

Summary

The Summary gives a simple analysis of each column with a graph.

For numerical columns, it counts the number of null values; and calculates the sum, the mean, the standard deviation, the min, the max, the lower and upper quantile, and the median.

For categorical columns, it counts the number of null values, of unique values, and the min and max length.

Association

The last part is the Association tab, which measures the relationship between two variables. The association between numerical variables is the correlation:

• a zero correlation indicates no relationship between the variables

• a correlation of +1 indicates a perfect positive correlation, meaning that as one variable goes up, the other goes up as well

• a correlation of –1 indicates a perfect negative correlation, meaning that as one variable goes up, the other goes down

If you need, you can use the More details button to better understand associations.

The Predict Cross Selling problem is a common challenge faced by businesses looking to maximize their sales opportunities by identifying additional products or services that a customer is likely to purchase. This predictive model falls under the multi-class classification category, where the objective is to predict the likelihood of a customer buying various products, based on their past purchasing behavior and other relevant data.

Dataset Essentials for Cross Selling

To effectively train a machine learning model for cross selling, you need a well-structured dataset that includes:

• Customer Demographics: Information like age, gender, and income, which can influence purchasing decisions.

• Purchase History: Detailed records of past purchases, indicating which products a customer has bought.

• Engagement Metrics: Data on customer interactions with marketing campaigns, website visits, and other engagement indicators.

• Product Details: Information about the products, such as category, price, and any special features.

A typical dataset might look like this:

Target Column: The Target_Product column is crucial as it represents the product that the model will predict the customer is most likely to purchase next.

Steps to Success with Graphite Note

1. Data Collection: Gather comprehensive, clean, and well-structured data.

2. Feature Selection: Identify the most relevant features that could influence the model's predictions.

3. Model Training: Utilize Graphite Note's intuitive platform to train your multi-class classification model.

4. Evaluation and Iteration: Continuously assess and refine the model for better accuracy and relevance.

The Advantage of Predict Cross Selling

• Enhanced Customer Experience: By understanding customer preferences, businesses can offer more personalized recommendations.

• Increased Sales Opportunities: Identifying potential cross-sell products can significantly boost sales.

• Data-Driven Decision Making: Removes guesswork from marketing and sales strategies, relying on data-driven insights.

• Accessibility: With Graphite Note, even non-technical users can build and deploy these models, making advanced analytics accessible to all.

In conclusion, the Predict Cross Selling model is a powerful tool in the arsenal of any business looking to enhance its sales strategy. With Graphite Note, this complex task becomes manageable, allowing businesses to leverage their data for maximum impact.

Media Mix Modeling (MMM) is a statistical analysis technique used to quantify the impact of various marketing channels on sales and other key performance indicators (KPIs). It helps businesses allocate their marketing budget more effectively by understanding the contribution of each channel to overall performance.

Dataset Essentials for Media Mix Modeling

A robust dataset for Media Mix Modeling should include:

• Time Period: The specific dates or periods for which the data is collected.

• Marketing Spend: The amount spent on each marketing channel during the period.

• Sales Data: The total sales achieved in the same time period.

• Channel Performance Metrics: Metrics like impressions, clicks, conversions, etc., for each channel.

• External Factors: Information on external factors like economic conditions, competitor activities, or seasonal events.

• Market Dynamics: Changes in market conditions, customer preferences, or product availability.

An example dataset for Media Mix Modeling might look like this:

Target Column: Totall Sales

Steps to Success with Graphite Note

1. Data Compilation: Gather comprehensive data across all marketing channels and corresponding sales data.

2. Model Development: Use Graphite Note, Regression Model, to develop a statistical model that correlates marketing spend across various channels with sales outcomes.

3. Analysis and Insights: Analyze the model's output to understand the effectiveness of each marketing channel.

4. Strategic Decision Making: Apply these insights to optimize future marketing spends and strategies.

Benefits of Media Mix Modeling

• Optimized Marketing Budget: Allocate marketing budgets more effectively across channels.

• ROI Analysis: Understand the return on investment for each marketing channel.

• Strategic Planning: Plan marketing strategies based on data-driven insights.

• Adaptability: Adjust marketing strategies in response to changing market conditions and consumer behaviors.

• Accessible Advanced Analytics: Graphite Note's no-code platform makes complex MMM accessible to teams without specialized statistical knowledge.

In summary, Media Mix Modeling is a powerful tool for businesses to optimize their marketing strategies based on comprehensive data analysis. With Graphite Note, this advanced capability becomes accessible, allowing for more informed and effective marketing budget allocation.

Customer Lifetime Value (CLV) prediction is a process used by businesses to estimate the total value a customer will bring to the company over their entire relationship. This prediction helps in making informed decisions about marketing, sales, and customer service strategies.

Dataset Essentials for Customer Lifetime Value Prediction

A suitable dataset for CLV prediction should include:

• Date: The date of each transaction or interaction with the customer.

• Customer ID: A unique identifier for each customer.

• Monetary Spent: The amount of money spent by the customer on each transaction.

An example dataset for Customer Lifetime Value prediction might look like this:

Steps to Success with Graphite Note

1. Data Collection: Compile transactional data including customer IDs and the amount spent.

2. Data Analysis: Use Graphite Note to analyze the data, focusing on customer purchase patterns and frequency.

3. Model Training: Train a model to predict the lifetime value of a customer based on their transaction history.

Benefits of Predicting Customer Lifetime Value

• Targeted Marketing: Focus marketing efforts on high-value customers.

• Customer Segmentation: Segment customers based on their predicted lifetime value.

• Resource Allocation: Allocate resources more effectively by focusing on retaining high-value customers.

• Personalized Customer Experience: Tailor customer experiences based on their predicted value to the business.

• Strategic Decision-Making: Make informed decisions about customer acquisition and retention strategies.

In summary, predicting Customer Lifetime Value is crucial for businesses to understand the long-term value of their customers. Graphite Note facilitates this process by providing a no-code platform for analyzing customer data and predicting their lifetime value, enabling businesses to make data-driven decisions in customer relationship management.

Predicting customer churn is a critical challenge for businesses aiming to retain their customers and reduce turnover. This problem typically involves a binary classification model, where the goal is to predict whether a customer is likely to leave or discontinue their use of a service or product in the near future.

Dataset Essentials for Customer Churn Prediction

A well-structured dataset is key to accurately predicting customer churn. Essential data elements include:

• Customer Demographics: Age, gender, and other demographic factors that might influence customer loyalty.

• Usage Patterns: Data on how frequently and in what manner customers use the product or service.

• Customer Service Interactions: Records of customer support interactions, complaints, and resolutions.

• Transaction History: Details of customer purchases, payment methods, and transaction frequency.

• Engagement Metrics: Measures of customer engagement, such as email opens, website visits, or app usage.

A typical dataset for churn prediction might look like this:

Target Column: The Churned column is the target variable, indicating whether the customer has churned (Yes) or not (No).

Steps to Success with Graphite Note

1. Data Gathering: Collect comprehensive and relevant customer data.

2. Feature Engineering: Identify and create features that are most indicative of churn.

3. Model Training: Use Graphite Note to train a binary classification model on your dataset.

4. Model Evaluation: Test the model's performance and refine it for better accuracy.

Benefits of Predicting Customer Churn

• Proactive Customer Retention: Identifying at-risk customers allows businesses to take proactive steps to retain them.

• Improved Customer Experience: Insights from churn prediction can guide improvements in products and services.

• Cost Efficiency: Retaining existing customers is often more cost-effective than acquiring new ones.

• Accessible Analytics: Graphite Note's no-code platform makes predictive analytics accessible, enabling businesses of all sizes to leverage AI for customer retention.

In summary, the Predict Customer Churn model is an invaluable tool for businesses focused on customer retention. Through Graphite Note, this advanced predictive capability becomes accessible to businesses without the need for extensive technical expertise, allowing them to make informed, data-driven decisions for customer retention strategies.

Expanding your dataset with more records can be beneficial for your machine learning model, but the impact depends on the quality and relevance of the new data. To learn how to re-upload or append data to your existing CSV dataset go .

Here’s how adding more records might help:

1. Improved Generalization

More data generally helps the model to learn better and generalize to unseen data. If your initial dataset was limited, your model might have overfitted to the specific patterns in that data. Adding more data helps the model capture a wider range of patterns, leading to better performance on new data.

2. Reducing Overfitting

When your dataset is small, the model may learn noise or irrelevant patterns (overfitting). Expanding the dataset introduces more variety, making it harder for the model to memorize specific samples, thereby helping to reduce overfitting.

3. Better Representing the Data Distribution

A larger dataset often better represents the underlying data distribution, especially if the new records cover more edge cases, outliers, or scenarios that were underrepresented in the original dataset. This helps the model become more robust and perform well across a wider range of inputs.

4. Enhanced Model Accuracy

In most cases, expanding your dataset improves the accuracy of the model, especially if the model is data-hungry (like deep learning models). More data means more examples for the model to learn from, allowing it to better predict future outcomes.

5. Handling Class Imbalance

If your dataset suffers from class imbalance (e.g., if one class has far more records than another in a classification problem), adding more records from the minority class can make your dataset more balanced, improving the model’s ability to predict minority classes correctly.

Considerations:

• Quality over Quantity: Simply adding more data isn’t always beneficial if the additional data is noisy, irrelevant, or incorrectly labeled. High-quality, representative data is more important than just increasing the size of the dataset.

• Data Diversity: Adding data that captures a wider variety of features or scenarios is more helpful than adding redundant or very similar data points. If the new data points are too similar to the existing ones, the impact on model performance might be minimal.

There are plenty of free sources to find free datasets for machine learning.

Here is a list of some of the most popular ones.

For each dataset, it is necessary to determine its quality. Several characteristics describe high-quality data, but it is essential to point out accuracy, reliability, and completeness. Every high-quality data should be precise and error-free. Otherwise, your data is misleading and inefficient. If your data is not complete, it is harder to use because of the lack of information. What if your data is ambiguous or vague? You cannot trust your data; it's unreliable.

Free datasets for machine learning

By googling stuff like free datasets for machine learning, time-series dataset, classification dataset, etc., you see many links to different sources. But which of them includes high-quality data? We will list a few sources, but it is essential to know that among them, there are also data that have their drawbacks. Therefore, you have to be familiar with the characteristics of a good dataset.

Kaggle

is a big data science competition platform for predictive modeling and analytics. There are plenty of datasets you can use to learn artificial intelligence and machine learning. Most of the data is accurate and referenced, so you can test or improve your skills or even work on projects that could help people.

Each dataset has its usability score and description. Within the dataset, there are various tabs such as Tasks, Code, Discussions, etc. Most datasets are related to different projects, so you can find other trained and tested models on the same datasets. On Kaggle, you can find a big community of data analysts, data scientists, and machine learning engineers who can evaluate your work and give you valuable tips for further development.

UCI Machine Learning Repository

is a database of high-quality and real-world datasets for machine learning algorithms. Datasets are well known in terms of exciting properties and expected good results; they can be an example of valuable baselines for comparisons. On the other hand, the datasets are small and already pre-processed.

GitHub

is one of the world’s largest communities of developers. The primary purpose of GitHub is to be a code repository service. In most cases within a project, we can find its application on some datasets; you will need to spend a little more time to find the wanted dataset, but it will be worth it.

data.world

Once you have found your dataset, it’s Graphite time; run several models and create various reports using visualizations and tables. With Graphite, it's easier to make business decisions. Maybe you are just a few clicks away from the turning point of your career.

Introduction

Data labeling is the process of tagging data with meaningful and informative labels to train machine learning models. In predictive analytics, labeled data is crucial as it provides the model with examples of correct behavior. This document will guide you through the process of preparing and labeling data for three predictive models:

Lead Scoring,

Churn Prediction,

and MQL to SQL Conversion.

1. Lead Scoring Model (Converted: Yes/No)

Objective: Predict if a lead will convert into a customer.

Dataset Example:

Steps:

1. Data Collection: Gather data on leads, including their industry, company size, and interactions with your platform.

2. Labeling: For each lead, label them as 'Yes' if they converted into a customer and 'No' if they didn't.

3. Reasoning: Labeling helps the model understand patterns of conversion based on the features provided.

2. Churn Prediction Model (Churned: Yes/No)

Objective: Predict if a customer will churn or leave your service.

Dataset Example:

Steps:

1. Data Collection: Gather data on customer usage patterns, support interactions, and feedback scores.

2. Labeling: For each customer, label them as 'Yes' if they churned and 'No' if they continued using your service.

3. Reasoning: Labeling helps the model identify signs of customer dissatisfaction or reduced engagement, which might lead to churn.

3. MQL to SQL Conversion Model (Converted: Yes/No)

Objective: Predict if a Marketing Qualified Lead (MQL) will become a Sales Qualified Lead (SQL).

Dataset Example:

Steps:

1. Data Collection: Gather data on MQLs, including their engagement with webinars, content downloads, and email interactions.

2. Labeling: For each MQL, label them as 'Yes' if they became an SQL and 'No' if they didn't.

3. Reasoning: Labeling helps the model recognize patterns of engagement that indicate a lead's readiness to move to the sales stage.

Conclusion

Data labeling is a foundational step in predictive analytics. By providing clear, accurate labels, you enable your predictive models to learn from past data and make accurate future predictions. Ensure your labels are consistent and based on well-defined criteria to achieve the best results with Graphite Note's no-code predictive analytics platform.

Overview

The MariaDB connector in Graphite Note allows you to import your data from a MariaDB or run custom SQL queries directly within the platform.

Prerequisites

Before starting, ensure your firewall allows incoming requests from the following IP address:

• 99.81.63.220

Steps to Import Data

1. Create a New Dataset

• Option 1: Go to the homepage and click on "Create" under Datasets.

• Option 2: From the datasets list, click on "New Dataset."

2. Select MariaDB

• Choose "MariaDB" as your dataset source and click "Next".

3. Enter Dataset Information

• Name: Provide a name for the dataset.

• Description: Add a short description of the data.

• Tags: Add tags for better organization.

• Click "Next" to proceed.

4. Establish a Connection

Fill in the following connection details:

• Server Name: Enter the hostname or IP address of your MariaDB instance.

• Database Port: Enter the port number for your database (typically 3306).

• Database User: Provide the username for the database.

• Database Password: Enter the password for the database user.

• Database Name: Specify the name of the database you wish to connect to.

• SSL (Secure Sockets Layer): SSL is a protocol for encrypting information over the internet. If your MariaDB instance requires SSL, ensure you enable this option.

5. Check the Connection

• Click on "Check Connection" to validate the connection details.

6. Write and Run SQL

• Write the desired SQL query to fetch your data.

• Click on the "Run SQL" button to execute the query and retrieve the data.

7. Review and Adjust Data

• You should see all the columns from the selected dataset appearing.

• If necessary, you can change column names, data types, or data formats.

8. Create the Dataset

• Click on the "Create" button to finalize and create your dataset.

Troubleshooting Connection Issues

Ensure your firewall settings are configured to accept incoming requests from the IP address mentioned above. This is crucial for establishing a successful connection between Graphite Note and your MariaDB Server.

Import Process

Once the connection is validated:

• Graphite Note will initiate the data import process.

• The duration of this process depends on the size of your dataset. Small datasets will import in a few minutes, while larger datasets may take longer.

Next Steps

Overview

The MySQL connector in Graphite Note allows you to import your data from a MySQL database or run custom SQL queries directly within the platform.

Prerequisites

Before starting, ensure your firewall allows incoming requests from the following IP address:

• 99.81.63.220

Steps to Import Data

1. Create a New Dataset

• Option 1: Go to the homepage and click on "Create" under Datasets.

• Option 2: From the datasets list, click on "New Dataset."

2. Select MySQL

• Choose "MySQL" as your dataset source and click "Next".

3. Enter Dataset Information

• Name: Provide a name for the dataset.

• Description: Add a short description of the data.

• Tags: Add tags for better organization.

• Click "Next" to proceed.

4. Establish a Connection

Fill in the following connection details:

• Server Name: Enter the hostname or IP address of your MySQL instance.

• Database Port: Enter the port number for your database (typically 3306).

• Database User: Provide the username for the database.

• Database Password: Enter the password for the database user.

• Database Name: Specify the name of the database you wish to connect to.

• SSL (Secure Sockets Layer): SSL is a protocol for encrypting information over the internet. If your MySQL instance requires SSL, ensure you enable this option.

5. Check the Connection

• Click on "Check Connection" to validate the connection details.

6. Write and Run SQL

• Write the desired SQL query to fetch your data.

• Click on the "Run SQL" button to execute the query and retrieve the data.

7. Review and Adjust Data

• You should see all the columns from the selected dataset appearing.

• If necessary, you can change column names, data types, or data formats.

8. Create the Dataset

• Click on the "Create" button to finalize and create your dataset.

Troubleshooting Connection Issues

Ensure your firewall settings are configured to accept incoming requests from the IP address mentioned above. This is crucial for establishing a successful connection between Graphite Note and your MySQL database.

Import Process

Once the connection is validated:

• Graphite Note will initiate the data import process.

• The duration of this process depends on the size of your dataset. Small datasets will import in a few minutes, while larger datasets may take longer.

Next Steps

If you've collected additional data related to your previously uploaded CSV file or there have been changes to the existing data, you can use the re-upload option. This allows you to update or append new data to your existing dataset. Expanding your dataset with more records can benefit your machine learning model, but the impact depends on the quality and relevance of the new data. Learn more about the benefits of dataset expansion .

How to Re-upload a CSV File:

1. Go to the Datasets List:

   • Navigate to the list of datasets.

1. Select the Dataset:

• Choose the dataset you wish to re-upload.

1. Select Re-upload:

• Click on the 'Re-upload' option.

1. Choose Options and upload:

• Decide on data append:

  • Depending on your needs, you can choose 'Append data' to add new records to the existing dataset. If 'Append data' is turned off, the new dataset will overwrite the existing

• Upload Your File:

  • Select or drag and drop your CSV file. Ensure the file has the same column structure as the previously uploaded file.

• Click Update to complete the re-upload process.

Overview

The Redshift connector in Graphite Note allows you to import your data from a Redshift database or run custom SQL queries directly within the platform.

Prerequisites

Before starting, ensure your firewall allows incoming requests from the following IP addresses:

• 99.81.63.220

Steps to Import Data

1. Create a New Dataset

• Option 1: Go to the homepage and click on "Create" under Datasets.

• Option 2: From the datasets list, click on "New Dataset."

2. Select Redshift

• Choose "Redshift" as your dataset source and click "Next".

3. Enter Dataset Information

• Name: Provide a name for the dataset.

• Description: Add a short description of the data.

• Tags: Add tags for better organization.

• Click "Next" to proceed.

4. Establish a Connection

Fill in the following connection details:

• Server Name: Enter the hostname or IP address of your Redshift instance.

• Database Port: Enter the port number for your database (typically 5432).

• Database User: Provide the username for the database.

• Database Password: Enter the password for the database user.

• Database Name: Specify the name of the database you wish to connect to.

• SSL (Secure Sockets Layer): SSL is a protocol for encrypting information over the internet. If your Redshift instance requires SSL, ensure you enable this option.

5. Check the Connection

• Click on "Check Connection" to validate the connection details.

6. Write and Run SQL

• Write the desired SQL query to fetch your data.

• Click on the "Run SQL" button to execute the query and retrieve the data.

7. Review and Adjust Data

• You should see all the columns from the selected dataset appearing.

• If necessary, you can change column names, data types, or data formats.

8. Create the Dataset

• Click on the "Create" button to finalize and create your dataset.

Troubleshooting Connection Issues

Ensure your firewall settings are configured to accept incoming requests from the IP addresses mentioned above. This is crucial for establishing a successful connection between Graphite Note and your Redshift data warehouse.

Import Process

Once the connection is validated:

• Graphite Note will initiate the data import process.

• The duration of this process depends on the size of your dataset. Small datasets will import in a few minutes, while larger datasets may take longer.

Next Steps

Oracle Connector for Graphite Note will soon be available for public use. This connector will enable seamless integration with Oracle databases, enhancing data accessibility and supporting advanced analysis within the Graphite Note platform. Stay tuned for its release to unlock efficient Oracle data connectivity.

In the next section, you’ll learn how to define a scenario, train a model, and leverage the results to make predictions, take strategic actions, and make data-driven decisions that directly impact your business.

Each model will be introduced on a dedicated page with step-by-step instructions and a video tutorial, guiding you through the process from setup to actionable insights.

In this section, we'll outline steps to enhance your model's performance.

The main idea behind Graphite Notebook is to do your own Data Storytelling; create various visualization with detailed descriptions, plot model results for better understanding, etc.

In the following sections, you will find more details about the Dataset API.

In the following sections, you will find more details about the Prediction API.

Request Payload

The request payload should be provided as the -d option in the cURL command. Replace [data] with the actual data you want to send for prediction. The structure and content of the data depend on the requirements of the specific model. More on this in next section.

Use this API to create new datasets for client onboarding, specifying the dataset's structure, settings, and initial data. This API is particularly useful for automating the setup of datasets during the onboarding process, allowing for easy integration with client-specific data requirements.

For example, use this endpoint to define the columns, types, and other properties for a new dataset tailored to a client's needs.

Create a Dataset

To create a new dataset, follow these steps:

1. To create a new dataset, make a POST request to /dataset-create with the required parameters in the request body.

2. To specify the dataset structure, include an array of column definitions with each column's name, alias, type, subtype, and optional format.

• Type can be:

  • measure

  • dimension

• Subtype can be:

  • text

  • numeric

  • date

  • datetime

1. To customize the dataset further, optional parameters such as CSV settings and a description can be included in the request.

The response will include key details about the created dataset, including the dataset code, table name, and the number of columns.

Example Usage

Creating a New Dataset

For example, making a POST request to the following URL with the provided JSON body would result in the response below: Request

Response

This request creates a dataset with the specified columns, each having unique names, types, and formats tailored to client onboarding requirements.

Overview

The MS SQL connector in Graphite Note allows you to import your data from a MS SQL or run custom SQL queries directly within the platform.

Prerequisites

Before starting, ensure your firewall allows incoming requests from the following IP address:

• 99.81.63.220

Steps to Import Data

1. Create a New Dataset

• Option 1: Go to the homepage and click on "Create" under Datasets.

• Option 2: From the datasets list, click on "New Dataset."

2. Select MS SQL

• Choose "MS SQL" as your dataset source and click "Next".

3. Enter Dataset Information

• Name: Provide a name for the dataset.

• Description: Add a short description of the data.

• Tags: Add tags for better organization.

• Click "Next" to proceed.

4. Establish a Connection

Fill in the following connection details:

• Server Name: Enter the hostname or IP address of your MS SQL instance.

• Database Port: Enter the port number for your database (typically 1433).

• Database User: Provide the username for the database.

• Database Password: Enter the password for the database user.

• Database Name: Specify the name of the database you wish to connect to.

• SSL (Secure Sockets Layer): SSL is a protocol for encrypting information over the internet. If your MS SQL instance requires SSL, ensure you enable this option.

5. Check the Connection

• Click on "Check Connection" to validate the connection details.

6. Write and Run SQL

• Write the desired SQL query to fetch your data.

• Click on the "Run SQL" button to execute the query and retrieve the data.

7. Review and Adjust Data

• You should see all the columns from the selected dataset appearing.

• If necessary, you can change column names, data types, or data formats.

8. Create the Dataset

• Click on the "Create" button to finalize and create your dataset.

Troubleshooting Connection Issues

Ensure your firewall settings are configured to accept incoming requests from the IP address mentioned above. This is crucial for establishing a successful connection between Graphite Note and your MS SQL Server.

Import Process

Once the connection is validated:

• Graphite Note will initiate the data import process.

• The duration of this process depends on the size of your dataset. Small datasets will import in a few minutes, while larger datasets may take longer.

Next Steps

• Now that your data is imported and prepared, you can proceed to create a model without writing any code. Simply go to the Models and follow the steps to build and deploy your model using Graphite Note's intuitive interface.

In Graphite Note, data preparation is divided into two main steps to ensure optimal results, with all tasks handled automatically so you don’t have to worry about them. Data preprocessing is a crucial step in machine learning, enhancing model accuracy and performance by transforming and cleaning the raw data to remove inconsistencies, handle missing values, and scale features, and ensure compatibility with the chosen algorithm.

Step 1: Exclusion of Columns

Features Not Fit for Model: Graphite automatically excludes columns that aren’t suitable for modeling, such as date/datetime columns, to ensure only relevant features are used in training.

Step 2: Preprocessing

To achieve the best results, Graphite Note takes care of several preprocessing steps:

• Null Values: It identifies and processes null values based on best practices. If the column is 50% null or more, the column will not be included in model training

• Missing Values: Missing values are managed automatically to maintain data integrity. For a numerical column it will change it by the average, and for a categorical feature it will become "not_available"

• One-Hot Encoding: Categorical variables are automatically transformed using one-hot encoding, converting categories into numerical formats suitable for model training.

• Fix Imbalance: Graphite addresses class imbalance in classification tasks, fixing the inequal distibution of target class and ensuring a balanced representation of classes.

• Normalization: Numeric columns are scaled to a uniform range, ensuring consistent data for models that require normalized input.

• Constants: Columns with constant values, which don’t contribute useful information, are identified and excluded from the dataset.

• Cardinality: Graphite optimizes high-cardinality categorical columns for model performance, handling complex categorical data effectively.

In traditional data science projects, these steps would require manual effort from data scientists, including data cleaning, encoding, scaling, and testing, often involving a significant amount of time and expertise. Graphite Note automates this entire process, completing these steps in seconds and allowing users to focus on insights and decision-making rather than data preparation.

Overview

Graphite Note’s CSV integration allows you to import your data from CSV files. This guide will walk you through the steps to import your CSV data into Graphite Note.

Steps to Import Data

1. Create a New Dataset

• Option 1: Go to the homepage and click on "Create" under Datasets.

• Option 2: From the datasets list, click on "New Dataset."

1. Select CSV file

• Choose "CSV" as your data source and click "Next".

1. Enter Dataset Information

• Name: Provide a name for the dataset.

• Description: Add a short description of the data.

• Tags: Add tags for better organization.

• Click "Next" to proceed.

1. Choose Parsing Options

• Configure your parsing options to properly interpret the CSV file.

  • Delimiter: Choose the character that separates values in your CSV (e.g., comma, semicolon).

  • Header: Indicate whether your CSV file has a header row.

  • Convert Empty Values to Null: Convert 'empty string' values to null. By default, this is turned on.

  • File Encoding: Select the file character encoding. The default is Unicode (UTF-8).

  • Skip Empty Rows: Skip empty rows at the beginning or end of the file. By default, Graphite Note will ignore those lines.

• Click on "Parse file" to process the file.

1. Review parsed data

   Review the parsed data and make any necessary adjustments:

   • Rename Columns: Change column names if needed.

   • Change Data Types: Adjust the data types of your columns as required.

   • Click on the "Create" button to finalize and create your dataset.

CSV filesize upload limits

Graphite Note limits the file size for CSV uploads to 50 MB. This restriction ensures that the platform maintains optimal performance, prevents excessive resource consumption, and ensures fast processing times for data uploads. If users need to work with larger datasets, alternative methods such as connecting directly to a database (Postgres, MySQL, BigQuery) can be used.

Next Steps

Now that your data is imported and prepared, you can proceed to create a model without writing any code. Simply go to the Models and follow the steps to build and deploy your model using Graphite Note's intuitive interface.

Overview

The PostgreSQL connector in Graphite Note allows you to import your data from a PostgreSQL database or run custom SQL queries directly within the platform.

Prerequisites

Before starting, ensure your firewall allows incoming requests from the following IP addresses:

• 99.81.63.220

Steps to Import Data

1. Create a New Dataset

• Option 1: Go to the homepage and click on "Create" under Datasets.

• Option 2: From the datasets list, click on "New Dataset."

2. Select PostgreSQL

• Choose "PostgreSQL" as your dataset source and click "Next".

3. Enter Dataset Information

• Name: Provide a name for the dataset.

• Description: Add a short description of the data.

• Tags: Add tags for better organization.

• Click "Next" to proceed.

4. Establish a Connection

Fill in the following connection details:

• Server Name: Enter the hostname or IP address of your PostgreSQL instance.

• Database Port: Enter the port number for your database (typically 5432).

• Database User: Provide the username for the database.

• Database Password: Enter the password for the database user.

• Database Name: Specify the name of the database you wish to connect to.

• SSL (Secure Sockets Layer): SSL is a protocol for encrypting information over the internet. If your PostgreSQL instance requires SSL, ensure you enable this option.

5. Check the Connection

• Click on "Check Connection" to validate the connection details.

6. Write and Run SQL

• Write the desired SQL query to fetch your data.

• Click on the "Run SQL" button to execute the query and retrieve the data.

7. Review and Adjust Data

• You should see all the columns from the selected dataset appearing.

• If necessary, you can change column names, data types, or data formats.

8. Create the Dataset

• Click on the "Create" button to finalize and create your dataset.

Troubleshooting Connection Issues

Ensure your firewall settings are configured to accept incoming requests from the IP addresses mentioned above. This is crucial for establishing a successful connection between Graphite Note and your PostgreSQL database.

Import Process

Once the connection is validated:

• Graphite Note will initiate the data import process.

• The duration of this process depends on the size of your dataset. Small datasets will import in a few minutes, while larger datasets may take longer.

Next Steps

• Now that your data is imported and prepared, you can proceed to create a model without writing any code. Simply go to the Models and follow the steps to build and deploy your model using Graphite Note's intuitive interface.

Model Scenario

A Timeseries Forecast Model is designed to predict future values by analyzing historical time-related data. To utilize this model, your dataset must include both time-based and numerical columns. In this tutorial, we'll cover the fundamentals of the Model Scenario to help you achieve optimal results.

For the Target Column, select a numeric value you want to predict. It's crucial to have values by day, week, or year. If some dates are repeated, you can aggregate them by taking their sum, average, etc.

Next, you can choose a Sequence Identifier Field to group fields and generate an independent time series and forecast forecast for each group. Keep in mind, these values shouldn't be unique; they must form a series and there is maximum of 500 unique values allowed as sequence identifier. If you don't want to generate independent time series for each group, you can leave this option empty.

Then, select the Time/Date Column, specifying the column containing time-related values. The Time Interval represents the data frequency—choose daily for daily data, yearly for annual data, etc. With Forecast Horizon, decide how many days, weeks, or years you want to predict from the last date in your dataset.

The model performs well with seasonal data patterns. If your data shows a linear growth trend, select "additive" for Seasonality Mode; for exponential growth, select "multiplicative." For example, if you see annual patterns, set Yearly Seasonality to True. (TIP: Plotting your data beforehand can help you understand these patterns.) If you're unsure, the model will attempt to detect seasonality automatically.

For daily or hourly intervals, you can access Advanced Parameters to add special dates, weekends, holidays, or limit the target value.

Advanced Parameters

We are constantly enhancing our platform with new features and improving existing models. For your daily data, we've introduced some new capabilities that can significantly boost forecast accuracy. Now, you can limit your target predictions, remove outliers, and include country holidays and special events.

To set prediction limits, enter the minimum and maximum values for your target variable. For example, if you're predicting daily temperatures and know the maximum is 40°C, enter that value to prevent the model from predicting higher temperatures. This helps the model recognize the appropriate range of the Target Column. Additionally, you can use the Remove Days of the Week feature to exclude certain days from your predictions.

Country holidays and special dates

We added parameters for country holidays and special dates to improve model accuracy. Large deviations can occur around holidays, where stores see more customers than usual. By informing the model about these holidays, you can achieve more balanced and accurate predictions. To add holidays in Graphite Note, navigate to the advanced section of the Model Scenario and select the relevant country or countries.

Similarly, you can add promotions or events that affect your data by enabling Add special dates option. Enter the promotion name, start date, duration, and future dates. This ensures the model accounts for these events in future predictions.

Combining these parameters provides more accurate results. The more information the model receives, the better the predictions.

Removing data points

In addition to adding holidays and special events, you can delete specific data points from your dataset. In Graphite Note, enter the start and end dates of the period you want to remove. For single-day periods, enter the same start and end date. You can remove multiple periods if necessary. Understanding your data and identifying outliers or irrelevant periods is crucial for accurate predictions. Removing these dates can help eliminate biases and improve model accuracy.

By following these steps, you can harness the full potential of your Timeseries Forecast Model, providing valuable insights and more accurate predictions for your business. Now it's your turn to do some modeling and explore your results!

Training model

After setting all parameters it is time to Run Scenario and train Machine Learning model.

The training duration may vary depending on the data volume, typically ranging from 1 to 10 minutes. The training will utilize 80% of the data to train various machine learning models and the remaining 20% to test these models and calculate relevant scores. Once completed, you will receive information about the best model based on the F1 value and details about training time.

Model Performance

To interpret the results after running your model, go to the Performance tab. Here, you can see the overall model performance post-training. Model evaluation metrics such as F1 Score, Accuracy, AUC, Precision, and Recall are displayed to assess the performance of classification models. details on Model metrics can also be found on Accuracy Overview tab.

On the performance tab, you can explore five different views that provide insights related to model training and results: Model Fit, Trend, Seasonality, Special Dates, and Details.

Model Fit

The Model Fit Tab displays a graph with actual and predicted values. The primary prediction is shown with a yellow line, and the uncertainty interval is illustrated with a yellow shaded area. This visualization helps assess the model's performance.

If you used the Sequence Identifier Field, you can choose which value to analyze in each Model Result.

Trend

Trends and seasonality are key characteristics of time-series data that should be analyzed. The Trend Tab displays a graph illustrating the global trend that Graphite Note has detected from your historical data.

Seasonality

Seasonality represents the repeating patterns or cycles of behavior over time. Depending on your Time Interval, you can find one or two graphs in the Seasonality Tab. For daily data, one graph shows weekly patterns, while the other shows yearly patterns. For weekly and monthly data, the graph highlights recurring patterns throughout the year.

Special Dates

The Special Dates graph shows the percentage effects of the special dates and holidays in historical and future data.

Details

Details tab shows the results of the predictive model, presented in a table format. Each record includes the predicted label, predicted probability, and predicted correctness, offering insights into the model's predictions, confidence, and accuracy for each data point. Dataset test results can be exporetd into Excel by clicking on the XLSX button in the right corner.

Take actions with Timeseries forecast

Once the model is trained, you can use it to predict future values, solve binary classification problems, and drive business decisions. Here are ways to take action with your Timeseries forecast model:

Predict

After building and analyzing a predictive model using Graphite Note, the "Predict" function allows you to apply the model to new data. This enables you to forecast outcomes or target variables based on different feature combinations, providing actionable insights for decision-making.

Create Notebook

You can share your prediction results with your team using the Notebook feature. With Notebooks, users can also run their own predictions on your Timerseries Forecast model.

Notebooks allow you to create various visualizations with detailed descriptions. You can plot model results for better understanding and enable users to make their own predictions. For more information, refer to the Data Storytelling section.

Graphite Note offers a suite of powerful machine learning and advanced analytics models designed to empower businesses to make data-driven decisions efficiently. Each model is tailored to address specific business needs, from forecasting future trends to segmenting customer bases. With these models, users can transform raw data into actionable insights quickly and without the need for complex coding.

Here’s a quick introduction to each type of model:

1. Timeseries Forecast: Ideal for predicting future values in timeseries data, such as sales or demand, based on historical patterns and seasonality.

2. Binary Classification: Used to classify data into two distinct groups (e.g., yes/no or true/false) based on historical data patterns.

3. Multi-Class Classification: Expands classification to multiple categories, allowing predictions across several possible outcomes.

4. Regression: A model that predicts a continuous numeric value (e.g., sales amount or customer age) based on other input features.

5. General Segmentation: Unsupervised learning that groups similar entities together, helpful in creating customer or product segments based on numeric similarities.

6. RFM Customer Segmentation: A specialized segmentation technique that segments customers based on Recency, Frequency, and Monetary value, aiding in targeted marketing.

7. Customer Lifetime Value: Predicts the future value of customers, estimating metrics like repeat purchase date and overall customer value for better retention strategies.

8. New vs Returning Customers: Provides insights into customer behavior by segmenting new and returning customers over various time frames (daily, weekly, monthly, etc.).

9. Customer Cohort Analysis: Groups customers based on their first purchase date, allowing businesses to analyze behavior patterns over time.

10. ABC Analysis: Categorizes items into A, B, and C categories based on their impact on a chosen metric, helping prioritize resources on high-impact items.

RFM Customer Model - How it Works?

Our intelligent system observes customers' shopping behavior without getting into the nitty-gritty technical details. It watches how recently each customer made a purchase, how often they come back, and how much they spend. The system notices patterns and groups customers accordingly.

This smart system doesn't need you to say, "Anyone who spends over $1000 is a champion." It figures out on its own who the champions are by comparing all the customers to one another.

When we talk about 'champion' customers in the context of RFM analysis, we're referring to those who are the most engaged, recent, and valuable. The system's approach to finding these champions is quite intuitive yet sophisticated.

Here's how it operates:

1. Observation: Just like a keen observer at a social event, the system starts by watching—collecting data on when each customer last made a purchase (Recency), how often they've made purchases over a certain period (Frequency), and how much they've spent in total (Monetary).

2. Comparison: Next, the system compares each customer to every other customer. It looks for natural groupings—clusters of customers who exhibit similar purchasing patterns. For example, it might notice a group of customers who shop frequently, no matter the amount they spend, and another group that makes less frequent but more high-value purchases.

3. Group Formation: Without being told what criteria to use, the system uses the data to form groups. Customers with the most recent purchases, highest frequency, and highest monetary value start to emerge as one group—these are your potential 'champions.' The system does this by measuring the 'distance' between customers in terms of RFM factors, grouping those who are closest together in their purchasing behavior.

4. Adjustment: The system then iterates, refining the groups by moving customers until the groups are as distinct and cohesive as possible. It's a process of adjustment and readjustment, seeking out the pattern that best fits the natural divisions in the data.

5. Finalization: Once the system settles on the best grouping, it has effectively ranked customers, identifying those who are the most valuable across all three RFM dimensions. These are your 'champions,' but the system also recognizes other groups, like new customers who've made a big initial purchase or long-time customers who buy less frequently but consistently.

By using this method, the system takes on the complex task of understanding the many ways customers can be valuable to a business. It provides a nuanced view that goes beyond simple categorizations, recognizing the diversity of customer value. The result is a highly tailored strategy for customer engagement that aligns perfectly with the actual behaviors observed, allowing businesses to interact more effectively with each segment, especially the 'champions' who drive a significant portion of revenue.

Why is ML better than rules-based segmentation?

Here’s why this machine learning approach is more powerful than manual labeling:

1. Adaptive Learning: The system continuously learns and adapts based on actual behavior, not on pre-set rules that might miss the nuances of how customers are interacting right now.

2. Time Efficiency: It saves you a mountain of time. No more going through lists of transactions manually to score each customer. The system does it instantly.

3. Personalized Grouping: Because it’s based on actual behavior, the system creates groups that are tailor-made for your specific customer base and business model, rather than relying on broad, one-size-fits-all categories.

4. Scalability: Whether you have a hundred customers or a million, this smart system can handle the job. Manual scoring becomes impractical as your customer base grows.

5. Unbiased Decisions: The system is objective, based purely on data. There’s no risk of human bias that might categorize customers based on assumptions or incomplete information.

In essence, this smart approach to customer grouping helps businesses focus their energy where it counts, creating a personalized experience for each customer, just like a thoughtful host at a party who knows exactly who likes what. It’s about making everyone feel special without having to ask them a single question.

RFM Model Scores

In the RFM model in Graphite Note, the intelligent system categorizes customers into segments based on their Recency (R), Frequency (F), and Monetary (M) values, assigning scores from 0 to 4 for each of these three dimensions. With five scoring options for each RFM category (including the '0' score), this creates a comprehensive grid of potential combinations—resulting in a total of 125 unique segments (5 options for R x 5 options for F x 5 options for M = 125 segments).

This segmentation allows for a high degree of specificity. Each customer falls into a segment that accurately reflects their interaction with the business. For example, a customer who recently made a purchase (high Recency), buys often (high Frequency), and spends a lot (high Monetary) could fall into a segment scored as 4-4-4. This would indicate a highly valuable 'champion' customer.

On the other hand, a customer who made a purchase a long time ago (low Recency), buys infrequently (low Frequency), but when they do buy, they spend a significant amount (high Monetary), might be scored as 0-0-4, placing them in a different segment that suggests a different engagement strategy.

By scoring customers on a scale from 0 to 4 across all three dimensions, the business can pinpoint exact customer profiles. This precision allows for highly tailored marketing strategies. For example, those in the highest scoring segments might receive exclusive offers as a reward for their loyalty, while those in segments with room for growth might be targeted with re-engagement campaigns.

The use of 125 segments ensures that the business can differentiate not just between generally good and poor customers, but between various shades of customer behavior, tailoring approaches to nurture the potential value of each unique segment. This granularity facilitates nuanced understanding and actionability for marketing, sales, and customer relationship management.

Model Scenario

Wouldn't be great to tailor your marketing strategy regarding identified groups of customers? That way, you can target each group with personalized offers, increase profit, improve unit economics, etc.

RFM Customer Segmentation Model identifies customers based on three key factors:

• Recency - how long it’s been since a customer bought something from you or visited your website

• Frequency - how often a customer buys from you, or how often he visits your website

• Monetary - the average spend of a customer per visit, or the overall transaction value in a given period

Let's go through the RFM analysis inside Graphite Note. The dataset on which you will run your RFM Model must contain a time-related column, given that this report studies customer behavior over some time.

We need to distinguish all customers, so we need an identifier variable like Customer ID.

If you might have data about Customer Names, great, if not, don't worry, just select the same column as in the Customer ID field.

Finally, we need to choose the numeric variable regard to which we will observe customer behavior, called Monetary (amount spent).

That's it, you are ready to run your first RFM Model.

RFM Model Results

As we now know how to run RFM model analysis in Graphite Note, let's go through the Model Results. The results consist of 7 tabs: RFM Scores, RFM Analysis, Recency, Frequency, Monetary, RFM Matrix, and Details Tabs. All results are visualized because a visual summary of information makes it easier to identify patterns than looking through thousands of rows.

RFM Scores

On the RFM Scores Tab, we have an overview of the customers and their scores:

Then you have a ranking of each RFM segment (125 of them) represented in a table.

And finally, a chart showing the number of customers per RFM score.

RFM Analysis