School of Mathematics and Statistics

Postgraduate research

Research being undertaken by our current postgraduates:

Contact

Michael Bertolacci


Supervisors

Start date

Mar 2016

Submission date

Links

Michael Bertolacci

Thesis

Bayesian mixture models for flexible modelling of longitudinal data

Summary

Bayesian mixture modelling is a statistical technique that assumes observations are generated from multiple probability distributions. Non-standard types of data can then be modelled by the ‘mixture’ of simple probability distributions. One class of mixture models are change point detection models for time series observations. These assume observations prior to a change point are generated by one probability distribution, observations post- change point from another, and that the exact point of change itself is unknown and must be estimated.

This project will extend existing mixture and change point methods to broad classes of longitudinal data, where longitudinal means time series measurements on collections of individuals. When there are many individuals and many observations per individual, and each individual has its own collection of multiple change points, the number of model parameters becomes very high.

Why my research is important

This project develops statistical and computational methods to handle issues that result from the combination of big data and complex statistical models. The techniques will be made available as software to the public, and will be sufficiently general to be used in a wide range of applications across the Engineering, Physical Sciences, Social Sciences, and Econometrics.



Statistics clinic

Assistance in statistics is available for Postgraduates students by research at the UWA Centre for Applied Statistics.

 

School of Mathematics and Statistics

This Page

Last updated:
Monday, 16 June, 2014 2:41 PM

http://www.maths.uwa.edu.au/275877