Write your own Bayesian Classifier: An Introduction to Machine Learning

Accepted Session
Long form
Scheduled: Thursday, June 18, 2009 from 1:45 – 3:30pm in Morrison


Can you perform simple arithmetic? Do you know how to program well enough to open and read files? Then you can write a Bayesian classifier, one of the machine learning techniques for predicting categories, most famous for its use in spam filters. Let's demystify this impressively-named but ultimately simple process.


Through the implementation of an honest-to-goodness Bayesian classifier, we’ll tour the major topics of supervised machine learning: tokenization, feature selection and vectorization, model training and tuning, and execution. Time permitting, we’ll touch on other techniques and topics.

Bring a laptop and an editor — at the end of the session, you should have your own classifier, understand how it works, and have some ideas for how to make it better.

Speaking experience