Association rule learning, or association rule mining, is a relatively modern unsupervised learning technique originally used to discover associations between purchased items in grocery stores. The goal of association rule mining is to discover interesting relationships between sets of items, for instance, discovering that shoppers preparing for a hurricane often buy Pop-Tarts along with their bottled water, batteries, and flashlights.
In Chapter 5, Classification Algorithms, we introduced the concept of conditional probability. In this chapter, we're going to take the concept a bit further and apply conditional probability to association rule learning. Recall that conditional probability asks (and answers) the question: given that we know something, what's the probability of something else happening? Or, what's the probability that someone...