Summary

In this chapter, we explored association rule learning, which is a branch of unsupervised learning. We implemented the Apriori algorithm, which can be used to find patterns in the form of rules in different transactional datasets. Apriori's classical use case is market basket analysis. However, it is also important conceptually, because rule learning algorithms bridge the gap between classical artificial intelligence approaches (logical programming, concept learning, searching graphs, and so on) and logic-based machine learning (decision trees).

In the following chapter, we're going to return to supervised learning, but this time we will switch our attention from non-parametric models, such as KNN and k-means, to parametric linear models. We will also discuss linear regression and the gradient descent optimization method.