Applications are invited for up to three PhD positions in ``Computational Bayesian Statistics with applications to decentralised machine learning’’ at the Department of Statistics, University of Warwick. The successful applicants will be part of the OCEAN project, a 10M Euro ERC-funded project involving around 30 researchers in Warwick, Paris and Berkeley. (Further details below.)  These studentships will be based at the University of Warwick Statistics Department which houses one of the largest and most successful groups in Computational Bayesian Statistics worldwide and successful applicants will join a vibrant local research community, as well as having the opportunity to interact closely with other researchers within the OCEAN consortium.

Applicants should have an interest and some background in Computational Bayesian methods interpreted widely, and applicants with more theoretical, more applied and/or more computational interests are equally encouraged to apply. Background in machine learning, privacy, game theory, federated learning and high-performance computing would be helpful for some topics but are not essential.

Successful applicants will have PhD fees covered from OCEAN as well as receiving an additional stipend beginning at £20,000 and access to a Research Training and Support fund of at least £5,000 to cover research equipment and travel costs.  Studentships will be for up to 4 years.

Informal enquiries before this deadline are strongly encouraged (to Gareth Roberts, Gareth.o.roberts@warwick.ac.uk, or Adam Johansen, a.m.johansen@warwick.ac.uk ). For more details on how to apply, please read on.


How to apply?
==============

Applicants should apply directly to the Warwick PhD admissions portal, all general information can be found at

https://warwick.ac.uk/study/postgraduate/apply/research

To be considered for these OCEAN positions, you are advised to apply before January 8th 2024 (after which we shall interview and make offers). (After this date, applicants are still welcome to apply but run the risk that positions will have already been filled.)

On your application it is important that you include the following:

1. You should apply to the Statistics Department.

2. Mention OCEAN explicitly in the personal statement part of the application. Please also mention it in the source of funding section. If you are potentially interested in other PhD opportunities in Warwick Statistics, you can indicate that also.

3. Also include in your personal statement information about your motivation and suitability for the OCEAN project. Precise information about an area you wish to work in is not necessary, although an indication about your research interests would be very helpful.

4. Please ensure that your referees are aware that they will need to upload their supporting statements by the deadline for these positions as decisions will be made soon after this deadline.

Competitive applications are likely to  require strong Undergraduate and Masters achievement, typically at first class and distinction levels respectively for UK applicants, and will most likely have completed degrees in Statistics, Mathematics, Data Science, Computer Science, although we are open to consider more unusual routes from motivated applicants.

Informal enquiries before this deadline are strongly encouraged (to Gareth Roberts, Gareth.o.roberts@warwick.ac.uk or Adam Johansen, a.m.johansen@warwick.ac.uk ). We'd be very pleased to hear from you and would be happy to advise about your application during a brief informal video call if that would be helpful.


The OCEAN project 
==================

Until recently, most of the major advances in machine learning and decision making have focused on a centralised paradigm in which data are aggregated at a central location to train models and/or decide on actions. This paradigm faces serious flaws in many real-world cases. In particular, centralised learning risks exposing user privacy, makes inefficient use of communication resources, creates data processing bottlenecks, and may lead to concentration of economic and political power. It thus appears most timely to develop the theory and practice of a new form of machine learning that targets heterogeneous, massively decentralised networks, involving self-interested agents who expect to receive value (or rewards, incentive) for their participation in data exchanges.

In response to these challenges, OCEAN  is an ERC-funded project which aims to develop statistical and algorithmic foundations for systems involving multiple incentive-driven learning and decision-making agents, including uncertainty quantification predominantly with a Bayesian focus. OCEAN will study the interaction of learning with market constraints (scarcity, fairness, privacy), connecting adaptive microeconomics and market-aware machine learning. To achieve these goals, OCEAN will need to develop new statistical and machine-learning methodologies, together with algorithms for sampling and optimisation which are both scalable to large problems, and have provable theoretical guarantees.

The OCEAN project is led by Eric Moulines (Ecole Polytechnique, Paris), Michael Jordan (Berkeley), Christian Robert (Dauphine, Paris) and Gareth Roberts (Warwick) and involves a consortium of around 30 researchers.