Study level

  • PhD
  • Master of Philosophy
  • Honours

Faculty/School

Faculty of Science

School of Mathematical Sciences

Topic status

We're looking for students to study this topic.

Research centre

Centre for Data Science

Supervisors

Dr Matt Sutton
Position
Lecturer in Statistical Inference for Complex Models
Division / Faculty
Faculty of Science

Overview

The goal of this project is to develop new continuous time Monte Carlo methods for efficient sampling from high-dimensional distributions. Continuous-time Monte Carlo methods are a class of algorithms that use continuous-time dynamics to generate samples from target distributions, rather than the discrete-time dynamics used in traditional Markov chain Monte Carlo (MCMC) methods. These methods have been shown to have faster mixing and better exploration of the state space, making them particularly appealing samplers for challenging distributions.

The main objectives of this project are to:

  1. Develop new continuous-time samplers for efficient sampling from complex distributions.
  2. Implement and test the proposed methods on a variety of challenging distributions, including but not limited to the ones arising in Bayesian statistics and machine learning.
  3. Compare the performance of the proposed methods with existing discrete-time MCMC methods.

Research activities

The student will:

  • have regular meetings to discuss the project and research topics
  • gain insights on developing new MCMC ideas
  • compare and analyse different sampling approaches.

Outcomes

Potential outcomes include:

  • code for continuous time Monte Carlo methods
  • comparison of samplers
  • potential coauthorship on a research article.

Skills and experience

This project is suitable for students with a background in mathematics, statistics, or computer science and a general interest in the field of probability and statistics. Familiarity with MCMC is not required.

Keywords

Contact

Contact the supervisor for more information.