ABSTRACT
This chapter discusses the different types of machine learning algorithms and some of the ways in which they are used. It explains what it means to build a model that will support augmented intelligence. Building models is complex because it requires that business leaders understand the business process they are focused on. There are two supervised learning algorithms that are most often used in model building—k-nearest neighbors and support vector machines (SVM). The chapter also discusses the process of building a model for a business problem from a machine learning algorithm running over well-prepared data. It explores several types of machine learning algorithms—inspectable ones, especially decision trees, and black box ones, including supervised and nonsupervised algorithms, reinforcement learning, and neural networks. There are many kinds of SVM algorithms, which grow in complexity with the mathematics that govern them. Unsupervised learning algorithms for clustering tasks fall into two broad classes: K-means clustering and hierarchical clustering.
