PhD Study : Active Bayesian machine learning

Summary

Overview

The aim of this project is to develop state-of-the-art Bayesian machine learning methods for use in applications where we generate our own training data using computer simulation. The research will be carried out within the Artificial Intelligence Research Group at Ulster University, and methods tested using a high impact application from the physical sciences, under active development at the University of Manchester.

Background

Recent failures in deep learning, such as the first fatality from driverless cars, have highlighted the importance of reporting the uncertainty in predictions made by machine learning [1]. The need to consider predictive uncertainty is common, and applies as much to the learning of atomic properties [2] as it does to autonomous driving. Bayesian machine learning [3] naturally provides us with estimates of the uncertainty in our predictions. The behaviour of a Bayesian model is a balance between what the training data says, and our pre-existing belief of how the model should behave, as expressed through probability distributions called priors.

This project concerns Bayesian machine learning for applications where we can generate our own training data using computer simulations. Choosing which data to learn from is called active learning, and is particularly interesting for Bayesian models given that the accuracy of our model depends also on the choice of prior.

Proposed research

The methods developed will be generic, but tested using a state-of-the-art application from computational chemistry that has been developed and updated over the past decade at the University of Manchester [4,5]. The exact research challenges to be addressed in this project can be tailored to the interests and experience of the PhD candidate, but example problems include:

*Active learning for multi-fidelity models: The simulations that generate the training data have tolerance parameters that enable us to control its fidelity (accuracy). Should we generate small amounts of high-fidelity data to train our models from, or large amounts of low-fidelity data, or a combination of both? How does this affect the uncertainty in our predictions?

*Prior specification: How can we best construct informative priors for a new problem, using (i) data from related problems, and (ii) domain knowledge, such as physical laws that place constraints on the data? How much data will we need to generate to counterbalance prior misspecification? Other challenges exist in areas such as dimensionality reduction, multi-task learning, and Bayesian optimization. All of these problems are of course not mutually exclusive, but overlapping.

[1] Kendall, A. (2018) Deep learning is not good enough, we need Bayesian Deep Learning for safe AI: Available at: https://alexgkendall.com/computer_vision/bayesian_deep_learning_for_safe_ai/

[2] Peterson, A., et al. (2017) Addressing uncertainty in atomistic machine leaning, Physical Chemistry Chemical Physics 19 10978-10985.

[3] Murphy, K. (2012) Machine Learning: A Probabilistic Perspective, MIT Press.

[4] Handley, C. M., Hawe, G. I., Kell, D. B., Popelier, P. (2009) Optimal construction of a fast and accurate polarisable water potential based on multipole moments trained by machine learning, Physical Chemistry Chemical Physics 11 6365-6376.

[5] McDonagh, J., et al. (2018) Machine Learning of Dynamic Electron Correlation Energies from Topological Atoms, Journal of Chemical Theory and Computation, 14 (1)

Essential criteria

Applicants should hold, or expect to obtain, a First or Upper Second Class Honours Degree in a subject relevant to the proposed area of study.

We may also consider applications from those who hold equivalent qualifications, for example, a Lower Second Class Honours Degree plus a Master’s Degree with Distinction.

In exceptional circumstances, the University may consider a portfolio of evidence from applicants who have appropriate professional experience which is equivalent to the learning outcomes of an Honours degree in lieu of academic qualifications.

Funding and eligibility

The University offers the following levels of support:

Vice Chancellors Research Studentship (VCRS)

The following scholarship options are available to applicants worldwide:

• Full Award: (full-time tuition fees + £19,000 (tbc))
• Part Award: (full-time tuition fees + £9,500)
• Fees Only Award: (full-time tuition fees)

These scholarships will cover full-time PhD tuition fees for three years (subject to satisfactory academic performance) and will provide a £900 per annum research training support grant (RTSG) to help support the PhD researcher.

Applicants who already hold a doctoral degree or who have been registered on a programme of research leading to the award of a doctoral degree on a full-time basis for more than one year (or part-time equivalent) are NOT eligible to apply for an award.

Please note: you will automatically be entered into the competition for the Full Award, unless you state otherwise in your application.

Department for the Economy (DFE)

The scholarship will cover tuition fees at the Home rate and a maintenance allowance of £19,000 (tbc) per annum for three years (subject to satisfactory academic performance).

This scholarship also comes with £900 per annum for three years as a research training support grant (RTSG) allocation to help support the PhD researcher.

• Candidates with pre-settled or settled status under the EU Settlement Scheme, who also satisfy a three year residency requirement in the UK prior to the start of the course for which a Studentship is held MAY receive a Studentship covering fees and maintenance.
• Republic of Ireland (ROI) nationals who satisfy three years’ residency in the UK prior to the start of the course MAY receive a Studentship covering fees and maintenance (ROI nationals don’t need to have pre-settled or settled status under the EU Settlement Scheme to qualify).
• Other non-ROI EU applicants are ‘International’ are not eligible for this source of funding.
• Applicants who already hold a doctoral degree or who have been registered on a programme of research leading to the award of a doctoral degree on a full-time basis for more than one year (or part-time equivalent) are NOT eligible to apply for an award.

Due consideration should be given to financing your studies. Further information on cost of living