Explain Your ML Predictions

AWS SageMaker Autopilot now allows users to concisely explain their models and machine learning model predictions from AWS SageMaker.

AWS SageMaker Autopilot allows users to automatically build, train, and tune machine learning models based on their data. This helps limit the heavy lifting of building ML models by only requiring that you provide a tabular dataset and select the target column to predict and SageMaker Autopilot will look for the best model to fit your data.

By better understanding how models make predictions and verifying that your model is behaving predictably, SakeMaker Autopilot helps making business decisions easier. And now more than ever with Model Exaplainability, SageMaker Autopilot provides, with help from Amazon SageMaker Clarify, a detailed report to understand and explain how the models you build with AWS SageMaker Autopilot make predictions. These reports include information about feature importance values which relate to each feature attribute and the percentage to which it contributes to the overall prediction.

The explainability report can be downloaded as a readable file, and helps stakeholders understand model characteristics as a whole prior to deployments.

Resources mentioned in this article:

• Amazon SageMaker
• Amazon SageMaker Autopilot
• Amazon SageMaker Clarify