ML governance
ML governance refers to everything that’s required to manage ML models within an organization, throughout the entire model development life cycle. As models can play a significant role in critical decision-making processes, it’s important to ensure that they are transparent, reliable, fair, and secure, and we need to implement structured frameworks to achieve these goals. These frameworks include policies and best practices that ensure responsible and ethical use of data and ML technologies.
When discussing ML governance in this chapter, I will also include data governance in the scope of the discussion, because the use of data is so inherent in the ML life cycle. Let’s start there.
Data governance
When it comes to managing data in the ML life cycle, there are a number of aspects that we need to consider, such as data quality, lineage, privacy, security, and retention. Let’s take a look at each of these in more detail.
