Financial Mathematics, Operational Research and Statistics Group
Finance, Operational Research and Statistics (FORS) group encompasses applied and theoretical research within three inter-related broad areas of research: financial mathematics, operational research and applied statistics.
The list below highlights a number of the research activities being undertaken, and lists the names of the various members of the group who are active in these areas. For more detailed descriptions of our research as well as other relevant information (collaborations, lists of publications, projects, conference presentations, and so on) please follow the links to the web pages of individual group members.
Financial Mathematics and operational research: The research in this area is focussed on financial modelling, derivative pricing and financial risk measurement and portfolio optimisation. The main topics of current research in financial mathematics include developing computationally efficient methods and models for forecasting the prices of financial derivative securities, efficient heuristics for risk measurement for derivative portfolios and the use of data analytics in risk measurement. The research is inherently multidisciplinary and uses tools from electrical engineering (particle filtering), operational research (mathematical programming) and applied statistics (efficient tail quantile estimation). Alumni of the research group work in several major financial institutions, including Blackrock, EY, Risk Care and Credit Suisse.
Statistics: Research in statistics spans a broad spectrum of methodology and applications. We are particularly strong in Bayesian methods for high-dimensional regression, clustering, variable selection, network analysis and quantile regression. Our methods are applied in bioinformatics, medicine, finance and engineering. Some of the ongoing research in statistics is being funded by external bodies such as the EU, the EPSRC and TWI.
Research interests and activities
- Nonlinear filtering and its applications in financial time series models (Date)
- High-dimensional multivariate statistical modelling, with applications on gene expression data (Vinciotti, Yu)
- Modelling of discrete data, with applications on reliability and deep-sequencing data (Vinciotti, Yu)
- Quantile regression and its applications in energy and environment ( Vinciotti, Yu)
- Applied Bayesian statistical inference (Vinciotti, Yu)
- Lifetime data analysis with application in engineering and health (Yu)
- Sampling from partially specified multivariate distributions, i.e. scenario generation (Date, Roman)
- Risk measurement and optimization of financial portfolios (Date, Lucas, Roman, Yu)
- Metaheuristic methods in combinatorial optimization (Beasley, Lucas)
- Social discounting, long term interest rate and actuarial mathematics (Meier)
Major international collaborators
- Prof Ernst Wit , Universita Svizzera Italiana, Lugano, Switzerland
- Dr Luigi Augugliaro, Dr Antonino Abbruzzo, University of Palermo, Italy
- Prof Peter-Bram ‘t Hoen, Radboud University Medical Center, Netherlands
- Dr Shovan Bhaumik, Indian Institute of Technology, Patna, India
- Dr Mohammad Hesamzadeh , KTH Stockholm, Sweden
- Dr Serdar Neslihanoglou, Eskehir Osmangazi University, Eskehir, Turkey
- Prof Rogemar Mamon, University of Western Ontarlo, London, Canada
- Dr Yanan Fan, UNSW, Sydney, Australia
- Prof Qifa Xu, Donghua University in Shanghai, China
- Prof Qifa Xu, Hefei University of Technology, China
- Prof Yan Yu, University of Cincinnati, USA
Externally funded projects
- Dr Vinciotti: short term scientific coordinator for an EU funded project COSTNET: the European Cooperation for Statistics of Network Data Science (2016-2020).
- Prof Yu: Three PhD studentships funded by TWI (Cambridge) on Statistical analysis and modelling of corrosion data, (2014—2020).
- Prof Yu: EPSRC Case Studentship: Data Analytics for Risk Based Decision Making in Asset Integrity Management (2018-2020).
- Dr Date: Industry-sponsored PhD studentship for research on the use of news analytics on volatility prediction (2014-2018).
Research opportunities for students
PhD applications are welcome in the following areas:
- Financial modelling; in particular, forecasting of spreads in commodity futures prices using latent state based models/ MCMC filters
- Applications of machine learning in financial models.
- Optimisation problems in power system transmission networks.
- Statistical modelling of multi-omics data. Specifically, prediction of proteomics data from transcriptomic data and identification of predictive regulatory motifs. This is in collaboration with Paola Vagnarelli, who will generate these data as part of her recently awarded Welcome Trust Grant.
- Computationally efficient correlated regression models. Specifically, the integration of efficient graphical modelling techniques within a regression framework. Potential application: credit risk modelling.
- Modelling the relationship or dependency as well as prediction among small variables with big or small data sizes via regression models and classification, such as quantile regression models.
- Statistical analysis of lifestyle interventions and economic evaluation for preventing Obesity, Asthma and other health issues
- Modelling paradigms and stochastic optimisation applied to (financial) decision making under uncertainty and risk