KobeShots

This repository contains an analysis of the Kobe Bryant Shot Selection dataset, using R 3.7.
The aim is to use different shallow learning techniques in order to perform a binary classification on the outcome of a shot. The algorithms considered are logistic regression, decision trees (as well as its variants, random forests and bagging), linear discriminant analysis, support vector machines and k-nearest neighbors.
In the case of logistic regression we also have a look a different techniques for model selection, such as subset selection and principal component analysis
Some of the ideas are taken from Aris Papadopoulos's article Machine Learning with the NBA's "Shooting Machine", which presents a similar approach in Python using decision trees.