50013

Machine Learning

Learning outcomes

By the end of the module you will be able to:
- Explain the strengths and weaknesses of machine learning algorithms.
- Appraise the suitability of a machine learning algorithm to solve a given problem.
- Provide appropriate methodologies to evaluate machine learning algorithms.
- Implement algorithms to solve machine learning problems.
- Develop predictive models with machine learning algorithms.
- Apply unsupervised clustering programs based on machine learning algorithms.

Module syllabus

This module covers the following topics:

1. Machine Learning Concepts: Definition and taxonomies, Supervised learning pipeline (classification/regression), Feature encoding, Model learning, Bias-variance tradeoff.
2. Instance Based Learning and Inductive Learning: K-nearest neighbours, Locally weighted regression, Decision trees.
3. Model Evaluation and Comparison: Training/validation/test data splits, Cross-validation, Evaluation metrics, Confidence intervals, Statistical significance.
4. Neural Networks: Perceptron, Multilayer perceptron, Back-propagation, Stochastic gradient descent, Activation and error functions, Overfitting, Data normalization, Hyper-parameter tuning.
5. Unsupervised Learning: Clustering algorithms (K-means), Models for density estimation (Kernel Density, Gaussian and Gaussian Mixture Models).
6. Evolutionary Algorithms: Genetic algorithms, Evolutionary strategies, Novelty search, Quality-Diversity optimisation.   

Teaching methods

The module is taught through lectures, practical exercises and quizzes to reinforce what has been taught in lectures, and computer-based coursework exercises done during lab sessions. The coursework exercises are in the form of mini-projects in which you will apply the methodology seen in class to solve machine learning problems.

An online service will be used as an open discussion forum for the module.

Assessments

There will be computer-based coursework exercises that contribute 30% of the marks for the module. In these coursework exercises, you will implement machine learning algorithms to solve practical problems. There will be a final written exam, which counts for the remaining 70% of the marks. Regular formative exercises will enable you to reinforce your understanding of the material taught in the class.

During the weekly live interactive sessions and lab sessions, teaching staff or assistants will be available to answer questions and give feedback on formative exercises. Detailed written feedback on the coursework exercises will be returned via department’s online platform.

Module leaders

Dr Marek Rei
Dr Josiah Wang
Professor Antoine Cully