A Leeds-Africa Collaborative Hub for Research in Data Analytics in Mathematical and Physical Sciences

The Hub’s Aims

This new Hub will drive collaborations between the UoL and less developed countries (LDCs) to tackle global challenges through physical and mathematical sciences. We will develop new collaborative research that addresses challenges in sustainable development, including climate science, bioinformatics, ecosystem functioning and industrial and technological advancement. By bringing together expertise at the UoL across (astro)physics, mathematics and statistics, along with the expertise and resources of overseas partners, this research agenda will enhance the development of LDCs while safeguarding both the local environment in these countries and across the world.

Planned Activities:

Workshops:

We plan our initial workshops to take place in South Africa during September 2023. The workshop will build connections with international partners and foster initial knowledge exchange. A possible subsequent workshop will be hosted at the University of Leeds.

These workshops will involve participants from a range of LDCs, with the hub funding the participation of researchers from LDCs. As part of these we will propose Hackathons on data science challenges related to exemplar themes with the Hub, such as single cell genomic data denoising and sparsity modelling (bioinformatics), rain data analysis and forecasting (climate science).

Research Exchanges:

A series of research exchanges, involving academics travelling both to and from the UoL for a period of c. 1 week each, and longer-term exchanges of PhD students, involving around 6 students in the first year traveling either from UoL to a partner country or vice versa, to foster deeper connections as part of long-term collaborations, and to broaden the horizons of these students. In the longer term we envisage joint supervision of master’s and PhD projects. This will help exploring new cutting-edge scientific problems, initiate new collaborations and consolidate existing ones with LDCs.

Long term:

A summer school in data analytics with applications to areas of mathematical and physical sciences in a partner country (likely the University of Pretoria) in the second year of this hub, which could involve around 50 students from both the UK and African countries, taught at a MSc/PhD level, involving tuition by academics at both UoL and the African partner institutions. As well as building research capacity for researchers in LDCs, this would provide a training in machine learning and data analysis for students from Leeds and (in limited numbers) elsewhere in the UK, as well as fostering the connections that could lead to joint PhD supervision and consolidation of research collaborations.

The long-term goal of the Hub is to foster lasting collaborations as a basis for externally funded research projects involving UoL and the partner institutions, which could take the form of research project grants or training networks/doctoral training centres.

Areas of anticipated research collaboration:

1. Machine learning for the characterisation and classification of objects and patterns in multi-wavelength imaging datasets for both observations and simulations.  Such tools would have wide ranging applications across areas such as Earth observation, medical imaging and astronomy.

2. Methods for data analysis in bioinformatics such as single cell genomic denoising and sparsity modelling. With continued technological progress in DNA sequencing, new forms of omics data are becoming available hence the call for novel modelling and statistical analysis to enrich the information obtained from sequencing experiments. This hub will contribute to the advancement towards currently unresolved challenges in topics such as trajectory analysis, pseudotime estimation, and network inference.

3. Analysis of animal tracking data and modelling of ecosystem function, in collaboration with faculty from the Department of Statistics. There is great interest in monitoring and tracking animal populations in African countries, with many European and American-based researchers carrying out fieldwork in, for example, Kenya, Tanzania and South Africa. This hub presents an opportunity to collaborate in the analysis of that data with African researchers and institutes

4. Applications of machine learning and AI in fluid dynamics. Leeds has a large research community in fluid dynamics centred at the Leeds Institute for Fluid Dynamics, including a Machine Learning/Data group with over 100 members. This presents an opportunity for a wide variety of collaborations, including projects in data analysis for fluids, as well as involvement of external partners in future bids for funding such as the AI Hub in Fluids (EPSRC).

5. Machine Learning and AI in Environment and Finance. This includes exploration of synergies between statistics and machine learning for the prediction of extremes in environmental problems like rainfall prediction, ML methods for insurance pricing and interpretability of ML/AI models.

Primary investigators for this project are: 

• Dr Richard Mann 
• Dr John Ilee