Computer Science

ECMM437 - Advanced Statistical Modelling (2019)

Back | Download as PDF
MODULE TITLEAdvanced Statistical Modelling CREDIT VALUE15
MODULE CODEECMM437 MODULE CONVENERUnknown
DURATION: TERM 1 2 3
DURATION: WEEKS
Number of Students Taking Module (anticipated) 20
DESCRIPTION - summary of the module content

***DATA SCIENCE AND DATA SCIENCE WITH BUSINESS STUDENTS ONLY***

The ideas of statistical modelling have been introduced in the two compulsory courses ECMM431From Data to Decisions and ECMM434 Machine Learning and Statistical Modelling. In this course we look at the concepts and methods of modern statistics in greater detail. The course will cover the philosophy and practice of Bayesian inference and how this differs from traditional methods of statistics. Bayesian hierarchical models are introduced as a method of acknowledging the inherent uncertainty that will be present in both data and the choice of statistical model. Methods are introduced for modelling structure within data, for example correlation over time and space, and for integrating data from multiple sources where the data collection mechanisms may differ. Bayesian methods require intensive computation, particularly for large datasets. This course covers modern computational statistical methods including Markov Chain Monte Carlo (MCMC) (including Hamiltonian MCMC), Approximate Bayesian Computation (ABC) and Integrated Nested Laplace pproximations (INLA).

Pre-requisites: ECMM431 From Data to Decisions and ECMM434 Machine Learning and Statistical Modelling
Co-requisites: None.

AIMS - intentions of the module

The aim of this module is to introduce you to modern methods in statistics, both conceptually and computationally, building on what you have learned in ECMM431 From Data to Decisions and ECMM434 Machine Learning and Statistical Modelling.

INTENDED LEARNING OUTCOMES (ILOs) (see assessment section below for how ILOs will be assessed)

On successful completion of this module you should be able to:

Module Specific Skills and Knowledge

1. Demonstrate a sound understanding of the ideas behind Bayesian statistics (and how it differs from classical or frequentist statistics).
2. Demonstrate the ability to carry out complex inferences on large datasets using modern statistical methods, such as MCMC, ABC and INLA, and to understand the underlying methodology to a level that enables modification of the computational approach to allow for non-standard problems and analyses.
3. Demonstrate the ability to model structure with data, for example correlation over space and time, and to be able to integrate data from multiple sources where data may be recorded at different points in space and time, and differences in data collection mechanisms may result in varying degrees of bias and uncertainty.

Discipline Specific Skills and Knowledge

4. Show sufficient knowledge of modern statistical methods both conceptual and computational.

Personal and Key Transferable / Employment Skills and Knowledge

5.  Reason using abstract ideas, formulate and solve problems and communicate reasoning and solutions effectively in writing.
6. Work effectively as part of a team.
7. Communicate orally with team members and via a poster and report.
8. Use learning resources appropriately.
9. Exhibit self management and time management skills.

 

SYLLABUS PLAN - summary of the structure and academic content of the module

Topics will include:
• The nature of probability
• Types of uncertainty
• Bayesian statistics and learning
• Bayesian hierarchical models
• Time series modelling
• Dynamic Linear Models
• Spatial models and the Gaussian Process
• Spatio-temporal models
• Analysing data in time and space
• Models for Data Integration
• Bayesian computation:
• Monte Carlo sampling
• Markov Chain Monte Carlo
• Gibbs Sampling
• Metropolis Hastings
• Hamiltonian MCMC
• Approximate Bayesian Computation
• Large-scale Bayesian computation
• Integrated Nested Laplace Approximations
• MCMC with the Stan language
• R-INLA

LEARNING AND TEACHING
LEARNING ACTIVITIES AND TEACHING METHODS (given in hours of study time)
Scheduled Learning & Teaching Activities 34.00 Guided Independent Study 114.00 Placement / Study Abroad 0.00
DETAILS OF LEARNING ACTIVITIES AND TEACHING METHODS
Category Hours of study time Description
Scheduled learning and teaching activities 18 Lectures
Scheduled learning and teaching activities 8 Practical classes in a computer lab
Scheduled learning and teaching activities 8 Tutorials
Guided independent study 116 Coursework preparation and background reading

 

ASSESSMENT
FORMATIVE ASSESSMENT - for feedback and development purposes; does not count towards module grade
Form of Assessment Size of Assessment (e.g. duration/length) ILOs Assessed Feedback Method
Workshop sheets 1h x 4 1-4 Feedback sheet
       
       
       
       

 

SUMMATIVE ASSESSMENT (% of credit)
Coursework 100 Written Exams 0 Practical Exams 0
DETAILS OF SUMMATIVE ASSESSMENT
Form of Assessment % of Credit Size of Assessment (e.g. duration/length) ILOs Assessed Feedback Method
Coursework report 100 2000-3000 words All Written
         
         
         

 

DETAILS OF RE-ASSESSMENT (where required by referral or deferral)
Original Form of Assessment Form of Re-assessment ILOs Re-assessed Time Scale for Re-assessment
Coursework report Coursework report All Within 8 weeks
       

 

RE-ASSESSMENT NOTES

Deferral – if you miss an assessment for certificated reasons judged acceptable by the Mitigation Committee, you will normally be either deferred in the assessment or an extension may be granted. The mark given for a reassessment taken as a result of deferral will not be capped and will be treated as it would be if it were your first attempt at the assessment.

Referral – if you have failed the module overall (i.e. a final overall module mark of less than 50%) you will be required to re-take some or all parts of the assessment, as decided by the Module Convenor. The final mark given for a module where re-assessment was taken as a result of referral will be capped at 50%.

RESOURCES
INDICATIVE LEARNING RESOURCES - The following list is offered as an indication of the type & level of
information that you are expected to consult. Further guidance will be provided by the Module Convener

Basic reading:

 

ELE: http://vle.exeter.ac.uk/

 

Web based and Electronic Resources:

 

Other Resources:

 

Reading list for this module:

Type Author Title Edition Publisher Year ISBN Search
Set Banerjee, S., Bradley, P. Carlin, A.& Gelfand, E. Hierarchical Modeling and Analysis for Spatial Data CRC Press 2014 [Library]
Set Gamerman, D. and Lopes H. F. Markov Chain Monte Carlo: Stochastic Simulation for Bayesian Inference CRC Press 2006 [Library]
Set Gelman, A., Carlin, J., Stern, H., Dunson, D., Vehtari, A. and Rubin, D. Bayesian data analysis 3rd CRC 2008 [Library]
Set Shaddick, G. & Zidek, J.V. Spatio-Temporal Methods in Environmental Epidemiology CRC Press 2015 [Library]
CREDIT VALUE 15 ECTS VALUE 7.5
PRE-REQUISITE MODULES ECMM431, ECMM434
CO-REQUISITE MODULES
NQF LEVEL (FHEQ) 7 AVAILABLE AS DISTANCE LEARNING No
ORIGIN DATE Tuesday 10 July 2018 LAST REVISION DATE Monday 14 January 2019
KEY WORDS SEARCH statistical modelling