Module Description

The aim of this module is to provide an understanding of the major approaches to machine learning, the methods involved in evaluating them, and their application to the solution of real problems.

LEARNING OUTCOMES

On completion of the course, students should be able to:

- demonstrate an understanding of the major approaches to classification and regression learning

- demonstrate an understanding of other machine learning techniques that have important practical applications

- identify machine learning techniques appropriate for particular classes of problem and apply them to practical problems

- undertake a comparative evaluation of several machine learning procedures

SYLLABUS

Introduction

- What is meant by machine learning

- Taxonomy of machine learning algorithms

- The inductive bias

- Data mining

Learning to classify:

- Decision tree induction

- Naïve Bayes methods

- Bayesian networks

- K-nearest neighbour method

- Support vector machines

Learning to predict numeric values:

- Linear Regression

- Regression trees

Evaluating learning procedures

Overfitting and the 'bias-variance trade-off'

Applications of machine learning

Clustering:

- k-means algorithm

- agglomerative hierarchical methods

Association rules mining:

- A priori algorithm

Reinforcement learning:

- Q learning

Multiple learners:

- Bagging, boosting, forests and stacking