# Unit information: Statistical Modelling in 2015/16

Please note: you are viewing unit and programme information for a past academic year. Please see the current academic year for up to date information.

Unit name Statistical Modelling MATHM6008 10 M/7 Academic Year (weeks 1 - 52) Dr. Kovac Not open None None School of Mathematics Faculty of Science

## Description including Unit Aims

This unit provides an introduction to many of the important aspects of statistical modelling, including the principles and practice of model selection, the role of conditional independence in modelling, techniques for graphical models, dealing with missing data and latent variables, and smoothing. The intention is to provide students with sufficient depth and generality of understanding that they can apply this knowledge across a broad range of modelling areas. The prerequisites are some previous exposure to linear and generalised linear models, likelihood theory and Bayesian inference.

Aims:

The aim of this course is to introduce important aspects of statistical modelling. Ideally, a broad range of types of model will also be incorporated into the course, but this is less essential. It is sufficient that the concepts below are covered in sufficient generality for students to be able to go on and address these issues in any model class with confidence. The overall objective is that students should acquire knowledge of important concepts of statistical modelling (as listed below), in sufficient depth and generality that they can apply this knowledge across a broad range of modelling areas.

Only available as part of a 1+ 3 Statistics MRes + PhD programme.

## Intended Learning Outcomes

For each of the following topics, students should understand:

• the issues (why this topic is important),
• the terminology,
• the statistical principles associated with this aspect of modelling,
• sufficient theory to deal with simple examples,
• and gain some practical hands-on experience in more complex examples.
1. Random effects/hierarchical/mixed models.
2. Principles and practice of model selection.
3. The role of conditional independence in modelling. Graphical models.
4. Missing data and latent variables.
5. Semiparametric models and smoothing

## Teaching Information

Lectures, supported by exercises, seminars and tutorials.

## Assessment Information

Assessment is though an extended assignment in the form of a "comprehension exercise". The students would have to read a recent paper from the literature involving advanced modelling, and to answer a series of pre-prepared questions, some of which may be reasonably open-ended.

The assessment criteria for the assignment will be based on a suitably modified version of the current Mathematics Department Project Assessment form. The assignment will be marked by the member of staff in charge of the unit and by an independent second marker.