Machine Learning. Feature Engineering. Principles and Techniques for Analysts

The authors of this book are machine learning specialist at Amazon, Alice Zheng, and analyst at the innovative IT company Concur Labs, Amanda Kazari. Feature engineering is the extraction of features from raw data and transforming them into a format suitable for processing by machine learning models. This is one of the most important processes in machine learning and simultaneously one of the most complex, as the diversity of models and data does not allow for a common feature engineering tactic to be identified. Nevertheless, the authors of the book have succeeded in this — deeper principles for working with data are formulated here, illustrated with specific examples. Each chapter describes the solution of various tasks: how to represent text data or images, how to reduce the dimensionality of automatically generated features, etc. In the last chapter, all examples are combined into a unified concept of feature engineering in machine learning. All code examples are provided in Python using modules such as NumPy, Pandas, Scikit-learn, Matplotlib, and are published in the authors' repository on GitHub.