Introduction

In the previous chapter, we learned various techniques for generating ensemble models by combining individual models. You will have noticed that building the ideal ensemble learning model involves a lot of experimentation with different base learners and meta learners. This is not the case for ensemble learning alone. The whole field of ML is all about performing various experiments to find the right combination of parameters and hyperparameters and enabling the extraction of performance from the models.

This process is a time-consuming one, with many different permutations and combinations that have to be tried before zeroing in on the ideal combination for a particular scenario. This is where the ML pipeline plays a big part. ML pipelines help in automating many of the tasks in the ML workflow. In this chapter, we will explore how ML pipelines can be used to automate ML workflows.

In the next section, we will define the business context before implementing the...