11. Machine learning#

In the previous two chapters, we built some intuition and skills around both statistical and machine learning (ML) approaches to making predictions about the world, and specifically for predicting numerical values (regression) and labels, or categories (classification).

In this chapter, we will:

  1. Think about when we might prefer to use a statistical versus machine learning approach (or both) to better understand or predict something.

  2. Consider some practical applications of regression vs. classification, and imagine how we might translate real world questions into regression or classification problems, including which type might be more useful (or even possible).

  3. Introduce unsupervised machine learning, including what it is, when we might use it, and get to know a simple yet powerful unsupervised ML algorithm called \(k\)-means.