Unit name | Stochastic Optimisation |
---|---|
Unit code | MATH30021 |
Credit points | 20 |
Level of study | H/6 |
Teaching block(s) |
Teaching Block 1 (weeks 1 - 12) |
Unit director | Dr. Tadic |
Open unit status | Not open |
Pre-requisites |
MATH11300 Probability 1 (or MATH10013 Probability and Statistics) and MATH20008 Probability 2 |
Co-requisites |
None |
School/department | School of Mathematics |
Faculty | Faculty of Science |
Unit Aims
The unit deals with the study of optimisation under uncertainty. It introduces some of the main modelling frameworks within which a wide variety of such problems can be set, before going on to study algorithms for their solution, and the analysis of these algorithms.
Unit Description
Stochastic optimisation covers a broad framework of problems at the interface of applied probability and optimisation. The unit will cover both static and dynamic problems. Static problems involve the optimisation of functions whose values are available only through noise-corrupted observations. Dynamic problems involve sequential decision-making to optimise some measure of long-term reward in a stochastic system evolving over time. The two main models studied in this context will be multi-armed bandit problems and Markov decision processes.
The unit will emphasise theoretical analysis of algorithms and derivation of optimal algorithms, as well as applications.
Students who successfully complete this unit should be able to:
Lectures, supported by problem and solution sheets.
Formative
10 problem sheets, approximately one per week, with feedback provided on selected problems. Full solutions will be provided for all problems.
Summative
100% Exam
Recommended