"Statistical thinking will one day be as necessary for efficient citizenship as the ability to read and write."
–H.G. Wells
In Chapter 3, Logistic Regression, we discussed using generalized linear models to determine the probability that a predicted observation belongs to a categorical response what we refer to as a classification problem. That was just the beginning of classification methods, with many techniques that we can use to try and improve our predictions.
In this chapter, we'll delve into two nonlinear techniques: K-Nearest Neighbors (KNN) and Support Vector Machines (SVMs). These techniques are more sophisticated than those we discussed earlier because the assumptions on linearity can be relaxed, which means a linear combination of the features to define the decision boundary isn't needed. Be forewarned, though, that this doesn't always equal superior predictive ability. Additionally, these models can be a bit problematic...