🐻 Lecture 23: Naïve Bayes

Yea yea… we have talked about estimating parameters for a parametric probability distribution. But now let’s actually do real ML: making predictions! We will do this in the context of three datasets: - Hearts - Ancestry - Netflix

All three problems on the pset where you implement classifiers for all of them using both Logistic regression and Bayes’ nets!

We will be working with supervised classification specifically.

Classification:

Note: regression is same thing except numbers aren’t binary!

So how do we program an algorithm for classification? Our first insight might be to brute force Bayes’.

Brute force Bayes’