We discussed supervised learning in detail in Chapter 2, Core Concepts in Machine Learning. Here, we are going to put into action some of the practices we learned. The most critical parts of any machine learning task are data exploration, cleaning, and feature representation. The process involves exploring the data, addressing the anomalies in the data, extracting features, feature selection, and feature reduction, if required. Almost 70 percent of the time in any data analytics project is spent in feature engineering, and it is the most important part of the analytical process.

Then comes the task of training models; selection of which machine learning algorithms to use is mostly guided by the available data and the objective of the problem we are about to solve.

Machine learning algorithms can be separated into two groups, based on the capability of a user to see how a model arrives at its predicted output. If we can deduce from the model how a particular prediction was done...