ST4MML: Methods of Machine Learning

Module code: ST4MML

Module provider: Mathematics and Statistics; School of Mathematical, Physical and Computational Sciences

Credits: 20

Level: 7

When you’ll be taught: Semester 1

Module convenor: Dr Fazil Baksh , email: m.f.baksh@reading.ac.uk

Module co-convenor: Dr Julia Abery, email: j.abery@reading.ac.uk

Pre-requisite module(s): BEFORE TAKING THIS MODULE YOU MUST TAKE MA1LA AND ( TAKE ST1PS OR TAKE MT2SWC ) (Compulsory)

Co-requisite module(s):

Pre-requisite or Co-requisite module(s):

Module(s) excluded: IN TAKING THIS MODULE YOU CANNOT TAKE ST3MML (Compulsory)

Placement information: NA

Academic year: 2025/6

Available to visiting students: Yes

Talis reading list: Yes

Last updated: 29 May 2025

Overview

Module aims and purpose

The topics of Data Science, Machine Learning and Artificial Intelligence are now part of the public consciousness, in part due to their successful application in industry. Many of the most successful techniques used in these fields are underpinned by statistical techniques. The aim of this module is to introduce students to a range of methods currently used in statistical machine learning, and to demonstrate how these are used in research and industry. The module begins by considering the application of unsupervised machine learning in exploratory multivariate data analysis and then moves on to consider methods of supervised machine learning in regression and classification. As well as being instructed in the theory underlying these methods, students will be given the opportunity to implement machine learning methods using statistical software and then to interpret and communicate their findings.

Module learning outcomes

By the end of the module, it is expected that students will be able to:

1. Use and explain a wide range of statistical methods used in Machine Learning, independently researching some methods beyond those taught, and.  justify which methods to use in different situations
2. Produce software implementation of the above methods and interpret findings
3. Communicate findings effectively to different audiences
4. Use statistical learning tools to build and evaluate algorithms for supervised learning

Module content

The module begins by considering the application of unsupervised machine learning in exploratory multivariate data analysis, covering the topics of data visualisation, principal component analysis, canonical variates analysis, cluster analysis and factor analysis. The module then discusses supervised machine learning, covering the topics of regression and classification, including: linear and logistic regression;  linear and quadratic discriminant analysis; resampling methods; model selection and regularisation; ridge regression; lasso; dimension reduction methods; principal components regression; partial least squares; high dimensional problems; regression splines; generalised additive models; tree-based methods; bagging; stacking; random forests; boosting; neural networks and deep learning; support vector machines.

Structure

Teaching and learning methods

The core material is delivered via lectures. These are supported by tutorials in which students work through non-assessed exercises and computer classes in which students practise the methods.

Study hours

At least 55 hours of scheduled teaching and learning activities will be delivered in person, with the remaining hours for scheduled and self-scheduled teaching and learning activities delivered either in person or online. You will receive further details about how these hours will be delivered before the start of the module.