Chapter 10: Understanding Model Results
In this chapter, you will learn how to analyze the results of your machine learning models to interpret why the model made the inference it did. Understanding why the model predicted a value is the key to avoiding black box model deployments and to be able to understand the limitations your model may have. In this chapter, you will learn about the available interpretation features of Azure Machine Learning and visualize the model explanation results. You will also learn how to analyze potential model errors and detect cohorts where the model is performing poorly. Finally, you will explore tools that will help you assess your model's fairness and allow you to mitigate potential issues.
In this chapter, we're going to cover the following topics:
- Creating responsible machine learning models
- Interpreting the predictions of the model
- Analyzing model errors
- Detecting potential model fairness issues
