In this taster week for MA335, we will be starting to look at feature or variable selection, specifically this is the problem of finding the best subset of features from a pool of available features in order to enhance our model's predictive ability. The importance of feature selection, what problems it tackles and some basic methodologies for variable selection are also discussed.
Firstly, you need to set aside a couple of hours and work through the videos and lecture notes below. Ideally, you should watch the videos and read the notes in the order given. Do make sure you take the time to read the notes in addition to watching the videos; you need to do both to make sure you have covered the materials in detail. To make this easier I have divided the materials below into three sections; Watch, Read, and Do. We have already discussed Watch and Read, the Do section is a short quiz to test what you have understood. I will be covering these materials with you on the the lecture / labs.
This is a sample week from the MA335: Modelling experimental and observational data module. Why don't you watch the lectures and try the exercises below?