CMPS142, Spring 2015, Section 01: Lecture Slides

Here are copies of the lecture slides for  CMPS 142, Spring 2015.

Notes:The second set of slides (LeastSq-Bias-Var) starts with the relationship between gauasian noise and least squares criterion (which we did in class on Tuesday 4/7), has a little bit on the closed form weights for linear regression (done differently via document camera and Andrew's notes), and ends with the bias-variance decomposition that we will do later.

The instance based slides were done Apr 30 before decision trees, and were gone over quickly, emphasizing the curse of dimensionality, the problem with irrelevant attributes, and the kernel-based density estimation.

Attachments