Dr Keming Yu
Tower A 025
- Email: firstname.lastname@example.org
- Tel: +44 (0)1895 266128
Current PhD projects:
- Bayesian quantile regression for Big Data.
- Development of machine learning methods for prediction.
- New distributional/regression methods for the analysis of Wellbeing, Covid-19 incubation.
- Statistical methods for substantive problems in health and biomedical scoences, such as obesity.
- Statistics/Machine learning methods for risk assessment in engineering, such as rail truck failure, cable fault, pipeline corrosion and wind turbine.
- Advanced statistical methods for financial econometrics, such as VaR.
Completed Research Students (2006--):
- Aseemali Fazal (MPhil), December 2006. Project Title: Dynamic Asset Pricing Using Statistics Forecasting Methods.
2. Yang Liu (MPhil), October 2007. Project Title: Modelling Credit Risk Securities.
3. Xiaochen Sun (PhD), May 2009. Project Title: Copula Methods for Risk Management in Finance.
4. Abdallah Ally (PhD), March 2011. Project Title: Quantile-based methods for prediction and risk analysis.
5. Craig Reed (PhD), April 2011. Project Title: ``Effective Gibbs sampling and variable selection for quantile regression with application‘’.
6. Antonios Koutsourelis (PhD), 2012. Project Title: ``Bayesian inference extreme quantile regression and hidden Markov models with application in risk analysis.’’
7. Balash Vladimir (MPhil), 2012. Project Title `` A new stochastic frontier production (SFP) model with application in econometric analysis of the efficiency and risk in Russian banking.’’
8. Zhuo Sheng (PhD), 2013. Project Title ``Extreme quantiles based volatility measurement in high-frequency data’’.
9. Rahim Al-Hamzawi (PhD), 2013. Project Title ``Prior elicitation and variable selection for Bayesian quantile regressions''. The winner of 2013 Vice-Chancellor's Prize for Doctoral Research.
10. Ali Al-Kenani (PhD), 2013. Project Title ``Dimension reduction in quantile regression’’.
11. Katerina Aristodemou (PhD), 2014. Project Title ``New Regression Techniques for Central Tendency’’.
12. Hussein Hashem (PhD), 2014, Project Title “Regularized and robust regression methods for high-dimensional data”.
13. Hamed Haselimashhadi (PhD), 2016, Project Title "Novel Regression Models for Dynamic and Discrete Response Data Under L1 and Differentiable Penalties" .
14. Hadeel kalktawi (PhD). 2017. Project Title ``Discrete Weibull Regression Model For Count Data’’.
15. Francisco Anes-Arteche (PhD), 2017 . Project Title `` Data Analysis for Improved Risk Assessment in Underground Pipelines’’. The winner of DEAN'S PRIZE FOR INNOVATION AND IMPACT.
16. Yang Hwei Tan (PhD), 2017, Project title: ``Statistical Methods for the Analyses of Corrosion Data for Integrity Assessments.
17. Xi, Liu, 2018 (PhD). Project title: `` Some New Developments for Quantile Regression’’.
18. Nik Nooruhafidzi Bin Muhd Noor (PhD), 2018. Project Title: `` Statistical modelling and prediction of defects of buried pipelines based on ECDA data and associated factors.’’
19. Eduardo Anes-Arteche (PhD) 2019-- `Risk assessment in Engineering'.
20. Sanna Soomro (PhD) 2020---`New models for spatial and high-dimensional adta analysis'.
21: Yuanqi Chu (PhD) 2020----`BQR for big data and COVID-19 data analsyis'.
PhD projects for research students
Wellbeing is taking centre stage in policy-making and the measurement of national success. Analysing wellbeing involves in challenges due to its measurement variety and choice as well somehow difficulty to explore the link between wellbeing and other factors. This project aims to explore and apply new regression models for wellbeing analysis, prediction and interpretation.
Weibull analysis is the popular method to make predictions about the life of all products in the population by fitting a statistical distribution to life data from a representative sample of units. Weibull analysis faces challenges from both small and big data. Moreover, the lifetime of a product may depend on many factors. This project aims to develop new Weibull distribution based methods or models for data analysis.
Landscapes are the result of numerous processes and factors that operate and interact across different spatial and temporal scales. These processes and factors include landscape structure process, social-economic factor, physical, biogeochemical, and anthropogenic factors. Regression models have been used in landscape modelling and ecological data analysis for decades but face new challenges. This project explores new spatiotemporal regression models to bring a new statistical method for decision making in landscape decision.
Extremes may give rise to bigger severity of occurrences than ordinary risks and then turn notably complex. Management of extreme events often faces challenges in the practice of problem solving and decision making, and therefore, requires of a special consideration and sufficiently using information provided. One of the most important types of statistical methods for gathering information for decision making is regression, when covariate information is available. How will the unprecedented data nowadays richness shape management of extreme events in practices? This projects aims to develop novel extreme-regression for risk analysis.