Introduction

Theory

An introduction to Generative Processes

Naive Bayes is usually the first time individuals are exposted to a generative process.

Statistical models like Least Squares and K nearest neighbors are discriminative models. A discriminative model is simply something that takes regressors $\vec{x}$ in order to predict a target value $y$.

Naive Bayes is a generative process. It differs in that it provides a join probability distribution [footnotes here explaining that].

Futher Resources

• The implentation of naive bayes in SKLearn is quite easy to understand and very well documented. It is so well written that I often use it for inspiration when writing my own analytics modules.