Summary

In this chapter, you learned about test-driven development using unit testing to control risks in your machine learning development projects. You learned about unit testing in Python using the pytest library. We also briefly reviewed the concept of differential testing, which helps you in comparing different versions of your machine learning modules and software. Later, you learned about model experiment tracking as an important tool that not only facilitates your model experimentations and selection but also helps you in risk control in your machine learning projects. You practiced using mlflow in Python as one of the widely used machine learning experiment tracking tools. Now, you know how to develop reliable models and programming modules through test-driven development and experiment tracking.

In the next chapter, you will learn about strategies to test models, assess their qualities, and monitor their performance in production. You will learn about practical methods...