About the Project: Bayesian Sampling has been very well studied in the literature for investigating various complex time series and regression analysis problems up to date. Due to their robust prior handling, and hierarchical sampling capabilities, methods such as Gibbs Sampling, and Markov chain Monte Carlo (MCMC) demonstrated sampling-based model analysis and prediction. However, due being random walk (RW) based their mixing capabilities and convergence are quite slow, and strongly depending on the proposal density defined prior to the implementations.
Non-reversible transitions, at this end, have started to play an important role and attracted significant attention in recent years and offer promising results thanks to their favourable convergence and mixing properties. The non-reversible transitions and their performance within Bayesian networks are still unexplored specifically for time-series prediction and trend tracking-like applications.
In this project, we will investigate exploiting superior mixing and convergence properties of non-reversible Bayesian sampling in complex network architectures. The developed algorithms will demonstrate BDL with faster convergence for various time series prediction examples whilst also leveraging the uncertainty quantification capabilities thanks to the utilised Bayesian framework. The initial application area will be environmental time series prediction of wind speed/power and other various meteorological time series data. Further, we exploit the trend tracking capabilities of the proposed approach with a potential application of target tracking in image sequences.
Keywords: Non-reversible sampling, Bayesian networks, Machine Learning, Time Series Prediction, Trend and Target Tracking.
Academic criteria: A 2:1 Honours undergraduate degree or a master's degree, in computing or a related subject. Applicants with appropriate professional experience are also considered.
Applicants for whom English is not their first language must demonstrate proficiency by obtaining an IELTS score of at least 6.5 overall, with a minimum of 6.0 in each skills component.
Desirable Criteria: Holding either a degree-level (applied-) statistics or (applied-) mathematics is a desirable criterion to conduct research in this project.
How to apply:
Please contact the supervisors of the project prior to submitting your application to discuss and develop an individual research proposal that builds on the information provided in this advert. Once you have developed the proposal with support from the supervisors, please submit your application following the instructions provided below
This project is accepting applications all year round, for self-funded candidates via https://www.cardiff.ac.uk/study/postgraduate/research/programmes/programme/computer-science-and-informatics
In order to be considered candidates must submit the following information:
- Supporting statement
- In the ‘Research Proposal’ section of the application enter the name of the project you are applying to and upload your Individual research proposal, as mentioned above in BOLD
- Qualification certificates and Transcripts
- Proof of Funding. For example, a letter of intent from your sponsor or confirmation of self-funded status (In the funding field of your application, insert Self-Funded)
- References x 2
- Proof of English language (if applicable)
If you have any questions or need more information, please contact COMSC-PGR@cardiff.ac.uk