Creating responsible machine learning models
Machine learning allows you to create models that can influence decisions and shape the future. With great power comes great responsibility, and this is where AI governance becomes a necessity, something commonly referred to as responsible AI principles and practices. Azure Machine Learning offers tools to support the responsible creation of AI under the following three pillars:
- Understand: Before publishing any machine learning model, you need to be able to interpret and explain the model's behavior. Moreover, you need to assess and mitigate potential model unfairness against specific cohorts. This chapter focuses on the tools that assist you in understanding your models.
- Protect: Here, you put mechanisms in place to protect people and their data. When training a model, data from real people is used. For example, in Chapter 8, Experimenting with Python Code, you trained a model on top of medical data from diabetic patients...
