Automatic Learning of Latent Force Models

There has been an increasing interest in Machine Learning to introduce inductive bias in predictive models through Physics. Such inductive bias allows a significant reduction in the number of data observations needed for training the model plus allows to increase the generalisation ability of the model at extrapolation tasks.

In latent force models, in particular, we combine differential equations with Gaussian processes by building kernel functions that encode dynamical systems. We have found applications of latent force models in a range of domains that include neural-engineering, robotics, computational biology and more recently in monitoring air pollution (see https://gow.epsrc.ukri.org/NGBOViewGrant.aspx?GrantRef=EP/T00343X/1
)

Building covariance functions for latent force models is usually a tedious process that involves analytically solving different integration problems. In this project, we will look at recent ideas developed in the kernel literature that automatically build covariance functions and apply and extend those approaches to the domain of latent force models.

This position will be open until filled with preferred starting date in April 2020.

The Supervisor

The candidate will be supervised by Dr Mauricio A Álvarez, Lecturer in Machine Learning at the Department of Computer Science, University of Sheffield. Dr Álvarez is well known for his work on Latent Force Models and multi-task learning using Gaussian Processes. More information can be found in the following link https://maalvarezl.github.io/

The Department

The Department of Computer Science at the University of Sheffield was established in 1982 and has since attained an international reputation for its research and teaching. In the latest Research Excellence Framework (REF2014), 45% of the research in the department was recognised as internationally excellent in terms of originality, significance and rigour, and another 47% as internationally world leading. These results place the department among the top 5 UK Computer Science departments for research excellence.

Candidate Requirements

The candidate is expected to have solid mathematical background and strong programming skills. Relevant experience and publications in the methods and/or applications above are preferred. These are in addition to the official requirements that must be satisfied (2nd upper/above, English). Please refer to the FAQ at https://www.sheffield.ac.uk/postgraduate/phd/research

A first degree in Mathematics, Statistics, Physics, Computer Science or Engineering.
An MSc in Mathematics, Statistics, Physics, Computer Science or Engineering with a 2.1 degree.
If English is not your first language, you must have an IELTS score of 6.5 overall, with no less than 6.0 in each component.

To apply for the studentship, applicants need to apply directly to the University of Sheffield using the online application system. Please name Mauricio Alvarez as your proposed supervisor.

Complete an application for admission to the standard Computer Science PhD programme
http://www.sheffield.ac.uk/postgraduate/research/apply

Applications should include a research proposal, CV, transcripts and two references.

The research proposal (up to 4 A4 pages, including references) should outline your reasons for applying for this scholarship and how you would approach the researching, including details of your skills and experience in Machine Learning