Skip to main content

Global Explanations (SHAP)

Welcome to the guide for generating global explanations for a model in Kranium AI. Global explanations provide insights into the overall behavior and decision-making patterns of your trained model. This guide will walk you through the steps to access the Explanation tab on the model details page and generate global explanations for your model.

1. Accessing the Model Details Page​

Steps:

  1. Navigate to the Models Section:
    • From the main dashboard, click on the "Models" tab to access the Models section.
  2. Select Your Trained Model:
    • In the Models listing screen, locate the trained model for which you want to generate global explanations and click on its name to open the model's details page.

2. Accessing the Explanation Tab​

Steps:

  1. Locate the Explanation Tab:
    • Within the model details page, find and click on the "Explanation" tab. This tab is typically located on the top of the page.
  2. Open the Explanation Tab:
    • Click on the "Explanation" tab to access the global explanation functionality. This tab provides tools and visualizations to understand the overall behavior of your model.

3. Generating Global Explanations​

Steps:

  1. Initiate Global Explanation:
    • In the Explanation tab, you will see options to generate global explanations for your model using SHAP (SHapley Additive exPlanations). Click on the "Generate Global Explanation" button to start the process.

4. Viewing and Analyzing Global Explanations​

Steps:

  1. Review Explanation Results:
    • The Explanation tab will display the global explanation results, providing insights into the factors that influence your model's overall behavior.
  2. Interpret Feature Contributions:
    • Examine the contribution of each feature to the model's predictions. This can help you understand which features are most important and how they impact the model's decisions.
  3. Analyze Model Behavior:
    • Analyze the global explanations to identify patterns and trends in the model's behavior. This can help you understand the model's strengths and weaknesses and identify areas for improvement.

5. Taking Action Based on Explanations​

Steps:

  1. Refine Model or Data:
    • Use the insights gained from the global explanations to refine your model or preprocess your data. This can help improve model performance and accuracy.
  2. Communicate Insights:
    • Share the explanation results with stakeholders to provide transparency and build trust in the model's predictions. Use visualizations and clear explanations to communicate the findings effectively.
  3. Monitor Model Performance:
    • Regularly generate and review global explanations to monitor the performance of your model over time. This can help you identify and address any changes in model behavior or performance.

Generating global explanations for your model in Kranium AI is a powerful way to gain insights into the overall behavior and decision-making patterns of your model. By following this guide, you can easily generate and analyze global explanations, helping you understand the factors influencing your model's outcomes and improve its performance. For any additional support or advanced configurations, refer to our support resources or contact the Kranium AI support team.