ABSTRACT
This chapter discusses ways to algorithmically tackle the problem of feature engineering using transformation functions in the context of supervised learning. It presents the techniques that focus on the feature construction aspects while utilizing feature selection as a black box. The chapter talks about general frameworks to automatically perform feature engineering in supervised learning through a set of transformation functions. It looks at the automation of the tasks described above for feature engineering in supervised learning using transformation functions. The chapter also looks at the strategies that automate the trial-and-error methodology, and those that try to learn patterns of association between features and effective transforms from past experience. Feature engineering is that task or process of altering the feature representation of a predictive modeling problem, in order to better fit a training algorithm. The hierarchical organization is a directed acyclic graph, known as a Transformation Graph.
