Summary
In this chapter, we established the framework for the different data processing units that will be introduced in this book. There is a very good reason why the topics of model validation and overfitting are treated early on in this book: there is no point in building models and selecting algorithms if we do not have a methodology to evaluate their relative merits.
In this chapter, you were introduced to the following topics:
The concept of monadic transformation for implicit and explicit models
The versatility and cleanness of the cake pattern and mixin composition in Scala as an effective scaffolding tool for data processing
A robust methodology to validate machine learning models
The challenge in fitting models to both training and real-world data
The next chapter will address the problem of overfitting by identifying outliers and reducing noise in data.
