Network on Computational Statistics and Machine Learning

The aim of this network is to establish the UK as the world leading authority in the joint area of Computational Statistics and Machine Learning (CompStat & ML) by advancing communication, interchange and collaboration within the UK between the disciplines of Computational Statistics (CompStat) and Machine Learning (ML).

The UK has tremendous research strength and depth that is widely acknowledged as world leading in both the individual areas of Computational Statistics and Machine Learning. Despite each of these fields of research developing, largely, independently and having their own separate journals, international societies, conferences and curricula both areas of investigation share a common theoretical foundation based on the underlying formal principles of mathematical statistics and statistical inference. As such there is a natural diffusion of concepts, research and individuals between both disciplines. This network will seek to formalise as well as enhance this interchange and in the process capitalise on important synergies that will emerge from the combined and shared research agendas of CompStat & ML.

Fit to EPSRC Strategic Priorities:

This proposed Research Network fits well within the EPSRC area of Statistics and Applied Probability identified as a priority 'grow' area for support. It meets squarely the stated aim of EPSRC to encourage 'greater connectivity with other research areas and facilitate multidisciplinary research'. As identified in the 2010 International Review in Mathematical Sciences there is fragility in the discipline of statistics in terms of a shortage of people. This proposed Network grant will be ideally suited to develop some of the most promising areas of research in statistics and assist in bringing on the new generation of young researchers required to grow Statistics in the UK.

Furthermore EPSRC has very recently identified that the "The interface between the mathematical sciences and ICT is extremely important and offers potential for high impact research. There are some well established connections between these disciplines but there are opportunities for new links to be developed." In the EPSRC ICT Theme, increasing the connections between Mathematical Sciences and ICT aligns to the Working Together cross-ICT priority and this Network proposal addresses this priority directly.


Communications and Engagement

The main ways in which it will be ensured that the beneficiaries can access the potential of the research output from this Network will be via the public access to materials through the dedicated website, further communication streams will be via Facebook and LinkedIn where professional contacts will seek specific research and technology expertise. The availability of the Research roadmap as it evolves will communicate the emerging research agenda to external agencies as well as direct adopters and developers of research results.

The annual Network workshops will be an excellent means with which to communicate developments within the Network and act as a means of establishing new collaborations within and without the Network to further widen access to Network activities.

In addition joint publications in the main journals and presentation at the main conferences of each of the respective areas covered will be pursued to ensure persistence of visibility of the outcome of Network based collaborations. Furthermore special sessions which will be jointly organised and supported by the network at the main leading international meetings (e.g. JSM, ISBA, NIPS, ICML) will provide an excellent way in which to propagate the joint research themes emerging from the Network.