The University of Southampton
Courses

COMP3206 Machine Learning

Module Overview

This module aims to introduce the mathematical foundations for machine learning and a set of representative approaches to address data-driven problem solving in computer science and artificial intelligence.

Aims and Objectives

Module Aims

This module aims to introduce the mathematical foundations for machine learning and a set of representative approaches to address data-driven problem solving in computer science and artificial intelligence.

Learning Outcomes

Knowledge and Understanding

Having successfully completed this module, you will be able to demonstrate knowledge and understanding of:

• Underlying mathematical principles from probability, linear algebra and optimisation
• The relationship between machine learning and neurophysiology
Subject Specific Practical Skills

Having successfully completed this module you will be able to:

• Systematically work with data to learn new patterns or concepts
• Gain facility in working with algorithms to handle data sets in a scientific computing environment
Subject Specific Intellectual and Research Skills

Having successfully completed this module you will be able to:

• Characterise data in terms of explanatory models
• Use data to reinforce one/few among many competing explanatory hypotheses

Syllabus

Historical Perspective - Biological motivations: the McCulloch and Pitts neuron, Hebbian learning. - Statistical motivations Theory - Generalisation: What is learning? - The power of machine learning methods: what is a learning algorithm? what can they do? Probability - Probability as representation of uncertainty in models and data - Bayes Theorem and its applications - Law of large numbers and the Gaussian distribution - Markov and graphical models Supervised Learning - Classification using Bayesian principles - Perceptron Learning - Support Vector Machines and Kernel methods - Neural networks/multi-layer perceptrons (MLP) - Features and discriminant analysis Linear Algebra - Using matrices to find solutions of linear equations - Properties of matrices and vector spaces - Eigenvalues, eigenvectors and singular value decomposition Data handling and unsupervised learning - Principal Components Analysis (PCA) - Blind source separation using Independent Components Analysis (ICA) - K-Means clustering - Spectral clustering - Manifold learning Regression and Model-fitting Techniques - Linear regression - Polynomial Fitting - Kernel Based Networks Optimisation - Convexity - 1-D minimisation - Gradient methods in higher dimensions - Constrained optimisation - Dynamic Programming Case Studies - Example applications: Speech, Vision, Natural Language, Bioinformatics.

Learning and Teaching

TypeHours
Follow-up work12
Revision10
Preparation for scheduled sessions12
Tutorial4
Supervised time in studio/workshop12
Lecture24
Total study time150

Mackay, David J. C.. Information Theory, Inference and Learning Algorithms.

Murphy, Kevin, (2012). Machine Learning: A Probabilistic Perspective.

James, G., Witten, D., Hastie, T., Tibshirani, R (2013). An Introduction to Statistical Learning: with Applications in R.

Bishop, Christopher M., (2006). Pattern Recognition and Machine Learning.

Barber, David (2012). Bayesian Reasoning and Machine Learning.

Assessment

Assessment Strategy

The coursework items will be varied in scope and will require different degrees of effort. Marks will be distributed accordingly, and the distribution will be made clear to students in advance.

Summative

MethodPercentage contribution
Coursework 30%
Coursework 20%
Exam  (2 hours) 50%

Referral

MethodPercentage contribution
Exam 100%

Repeat Information

Repeat type: Internal & External