Sunday, November 14, 2010

Statistical Learning Assignment 2

This post is about assignment 2 of a statistical learning master course. The goal here is to fit a Logistic regression and Naive Bayes model on two training/testing sets. the UCI Adult set and Voting dataset. The results were that for small training sets, Naive Bayes works better, and for large datasets, it's Logistic Regression that is better. It's a classic tale of Bias and Variance playing a role however, which is quite predictable. The report can be found on the below link:
https://docs.google.com/viewer?a=v&pid=explorer&chrome=true&srcid=0B9QrpxMuvSbnZDU3OGNkY2ItNmVlNC00MTJjLWEwYzMtMWQzMGNhZDE3NDA4&hl=en

No comments: