Introduction to machine learning

This unit develops your ability to model multi-dimensional data using statistical and machine learning techniques. Topics covered include: dimension reduction with linear and nonlinear methods; supervised learning such as discriminant analysis, decision trees and forests, neural networks; and unsupervised learning such as k-means, hierarchical and model-based clustering. You will learn about conceptualising problems using the bias-variance trade-off and how to balance this when fitting models. Complex model fitting techniques will be covered including bagging, boosting, cross-validation, regularisation and constructing ensembles. An important component is learning how to diagnose your model, especially utilising high-dimensional visualisation methods, and explain your model with explainable artificial intelligence (XAI). You will develop practical skills in applying techniques to different problems using a suitable software environment that involves doing reproducible analyses.