Foundation concepts you need for this module
Before starting this module, you may want to revise the optional foundation material on Probability again. In particular, you need to understand the properties of Gaussian probability density functions, and why modelling covariance requires so many parameters. This provides the motivation for the step in feature engineering where we try to eliminate (or substantially reduce) covariance.
If you are unsure what the logarithm is, then you should go through all the “Logarithms” videos in this foundation material on mathematics.
Module content
This module introduces the concept of feature engineering, in which the features are manipulated to make them more suitable for whichever type of machine learning we have decided to use.
Here’s what you’re going to learn in this module:
Lecture Slides
Slides for Thursday lecture (google)