module specification

This module will introduce students to modern statistical modelling techniques and how those techniques  can be used for prediction and forecasting. Throughout the statistical environment  and software R will be used in conjunction with relevant statistical libraries.
The module will,  introduce modern regression techniques (including smoothing),  discuss different model selection techniques (including the classical statistical hypothesis) and  how those techniques can be used for prediction purpose. 

Prior learning: Statistical knowledge desirable.

The module aims to:

1. Equip graduate students with modern statistical techniques
2. Provide students with some selected advanced statistics topics including forecasting 
3. Prepare students to be able to read and understand professional articles
4. Prepare students to carry on their own research and use modern statistical techniques as one of the tools for their research.

1. Introduction to probability, distribution functions and inference
           Different types of probability distribution functions
           Statistical inference and hypothesis testing, (LO1, LO3)

2. Flexible regression models,  for continuous and categorical data including
         i) Generalised Linear Models (GLM)
         ii) Generalised Additive Models (GAM) and
         iii) Generalised Additive Models for Location Scale and Shape (GAMLSS)      
     Introduction to smoothing techniques. (LO2, LO5, LO6, LO7)

3. Model  selection techniques and forecasting:
        i) Likelihood ratio test 
        ii) Generalised Akaike information criterion
        iii) The bias versus variance dilemma 
        iv) Forward and backward and stepwise selection techniques
        v) ridge regression and lasso
        v) Cross validation techniques 
        vi) Validation and test samples techniques
        vi) Boosting  (LO3, LO4, LO5, LO6, LO7)

The module will be delivered through a combination of lectures and associated tutorial and laboratory workshops over a period of 12 weeks. Topics of lectures will be supplemented with laboratory sessions to illustrate the application of the techniques studied. The R software will be used and students will be encouraged to broaden their knowledge by exploring complex real world data sets. Critical evaluation of the techniques used will be encouraged. The tutorial and lab sessions will also provide opportunities for students to obtain informal feedback from the teaching staff on their progress.

Additional teaching and learning resources will be made available via WebLearn and students will be expected to spend a significant proportion of their time on private study.

On completing the module, students will be able to:

[LO1] Demonstrate comprehensive knowledge and understanding of modern statistical modelling techniques, and their relation to traditional statistical analyses
[LO2] Use R to analyse a wide range of univariate response data, with
     explanatory variates, factors and smoothing
[LO3] Gain a considerable exposure to build a statistical model and evalute outcomes against its assumptions
[LO4] Appraise and assess the power and limitation of the prediction (forecasting) techniques
[LO5] Carry out independent investigation, research and write clear and concise scientific reports, partly through a peer review process.
[LO6] Demonstrate an excellent understanding of some current developments in statistical modelling and forecasting.
[LO7] A critical evaluation and a clear understanding of the applications of legal, social, ethical and professional issues to academic research and PhD programmes.

https://londonmet.rl.talis.com/modules/ma7007

Aitkin M. Francis B. Hinde J. Darnell R. (2009) Statistical modelling in R, Oxford Statistical Science Series 

Hastie T, Tibshirani  R. and Friedman J. (2009) The Elements of statistical Learning. Data Mining, Inference and Prediction. Second Edition, Springer [CORE]

Stasinopoulos D.M. Rigby, R.A. Voudouris V, Heller G. De Bastiani F. (2015) Flexible Regression and Smoothing, The GAMLSS packages in R. First Draft

http://www.gamlss.org/wpcontent/uploads/2015/07/FlexibleRegressionAndSmoothingDraft-1.pdf

Rigby R. A., Stasinopoulos D. M., Heller G., and De Bastiani. F. (2019). Distributions for Modelling Location, Scale and Shape: Using GAMLSS in R. Chapman and Hall/CRC,