Week 2 to 4
Instructor: Dr Keegan Kang
Description:
The three-week SHARP course will be on computational statistics, and will comprise a theoretical and computational component. The theoretical component consists of a brief review of A-Level statistics, before moving onto variance reduction techniques and an introduction to Monte Carlo methods. The practical component involves using Matlab, understanding (some) limitations of scripting languages, plotting basic graphs to compare certain quantities of interest, and (how to) critique them as well - amidst the backdrop of the theoretical portion. Lecture notes and code will be provided.
Students are to download and install Matlab before the first class - instructions will be provided in the lecture notes.
Assessment:
Assessment will be based on homework (done in pairs) and exam.
Brief outline of topics covered: (subject to slight changes)
Class 1: (half theory, half Matlab)
- Introduction
- Review of A-Level statistics
- Difference between theoretical / empirical expectation and variance and why it is important
- Examples of some of these differences used in practical situations
Class 2: (mostly Matlab)
- Limitations of Matlab (and other programming languages) ; numerical plots, experiments
- Vectorisation, more examples of plotting / visualising data
Class 3: (mostly theory with the use of Matlab to verify)
- Maximum likelihood estimation
- Monte Carlo integration
- Variance reduction techniques
- Brief spiel on research