banner-city

PhD in Statistics

Study with leading statisticians at a world-class university

Applications for entry 2021/22 are open

Apply here

Funding deadlines: 14 January and 29 April 2021.
Final application deadline: 27 May 2021 

A PhD offers the chance to undertake a substantial piece of supervised work that is worthy of publication and which makes an original contribution to knowledge in a particular field. Our PhD programme is designed to produce professional social scientists, well versed in a range of advanced statistical techniques and methods, in addition to having an in-depth knowledge of your topic of interest. 

The Department of Statistics is one of the world's leading centres of quantitative methods in the social sciences and has long been home to some of the world's most famous and innovative statisticians. Today, the department has an international reputation for the development of statistical methodology that has grown from our long history of active contributions to research and teaching in statistics. 

Our core research areas are:

  • Data science
  • Probability in finance and insurance
  • Social statistics
  • Time series and statistical learning

If you have any questions about our MPhil/PhD Statistics programme, please email our research manager

 

Research environment

The Department of Statistics at LSE is one of the oldest and most distinguished in the UK. It has a rich research portfolio covering core areas of statistical inference and real applications, particularly in the economic, financial and actuarial, social and industrial arenas. The close collaboration with other LSE departments, our London location and strong international partnerships are reflected in the research life of the Department of Statistics through the members of staff, PhD students, postdoctoral research fellows and the thriving visitor and seminar programmes.

Research in the department is concentrated in the following areas and PhD proposals should normally be linked to one of these areas:

Data Science

Research in the data science area is concerned with the development of new machine learning and statistical methods, and their applications. The areas of applications include the design of novel methods for understanding user behaviour, analysis of social data, modelling and inference for information cascades and epidemic processes that arise in social networks and biomedical applications, as well as algorithms for development of next-generation artificial intelligence systems.

Possible areas of research include:

  • Bayesian inference and predictions
  • Functional data analysis
  • High-dimensional statistics
  • Machine and statistical learning for relational data
  • Network data models, inference and predictions
  • Optimisation and machine learning
  • Reinforcement learning
  • Statistical learning methods in precision medicine
  • Statistical models and inference for ranking data
  • Stochastic models of epidemic processes
  • Stochastic optimisation methods
  • Stochastic processes in econometrics and finance

For more information about potential supervisors and their areas of interest, visit the Data Science research group.

Probability in Finance and Insurance

PhD research in probability in finance and insurance encompasses many aspects of the discipline. Methodological and theoretical research is mainly guided by applications with the aid of both academic and industrial experts, covering topics of modern stochastic finance with an emphasis on insurance and financial mathematics.  Applications include pricing and hedging exotic products, counterparty risk, portfolio optimisation, risk management and insurance, risk transfer and securitisation, etc. 

Research topics may be identified in advance by the applicant or may be arrived at through communication with a potential supervisor. The relative emphasis on methodology/theory vs. application may vary. 

For more details about supervisors and their areas of research interests, please see the Probability in Finance and Insurance research group. You will find links to the web pages of individual members of staff here. If you are interested in applying to undertake PhD research in probability in finance and insurance, you are welcome to contact one of these members of staff regarding a suitable topic for your research proposal. 

Suggested research areas of PhD research projects include:

  • Energy markets
  • Excursions of Lévy processes and applications in finance and insurance
  • Financial market with frictions
  • Information asymmetry
  • Interface between insurance and finance
  • Lévy processes
  • Optimal stopping
  • Point processes in insurance and credit risk
  • Quantile options and options based on occupation times
  • Stochastic analysis and its applications in financial mathematics
  • Stochastic control and analysis of partial differential equations in mathematical finance

This list is indicative only and by no means exhaustive.

Social Statistics

PhD programmes of study in social statistics typically include both methodological development and the application of statistical methods to a social science field or to address new developments in social data, such as in sample surveys or social networks. Research topics may be identified in advance by the applicant or may be arrived at through communication with a potential supervisor. The relative emphasis on methodology/theory vs. application may vary. 

Possible areas of research include:

  • Analysis of complex survey data
  • Disclosure risk assessment and statistical disclosure control
  • Estimation from survey data (and related data), taking account of nonresponse and using auxiliary information
  • Latent transition and latent class models for modelling diagnostic tests
  • Latent variable models and structural equation models for categorical data
  • Longitudinal data analysis, especially event history (survival) analysis and dynamic panel models
  • Modelling response strategies and detection of outliers in educational and behavioural sciences
  • Multilevel simultaneous equations modelling of correlated social processes

For more details about potential supervisors and their areas of interest, visit the Social Statistics research group. If you are interested in applying to undertake PhD research in social statistics, you are welcome to contact one of these members of staff regarding a suitable topic for your research proposal.

Time Series and Statistical Learning

PhD research in time series and statistical learning encompasses many aspects of these disciplines. We are keenly involved in both theoretical developments and practical applications. Current areas of interest include time series (including high-dimensional and non-stationary time series), data science and machine learning, networks (including dynamical networks), high-dimensional inference and dimension reduction, statistical methods for ranking data, spatio-temporal processes, functional data analysis, shape-constrained estimation, multiscale modelling and estimation and change-point detection.

Research topics may be identified in advance by the applicant or may be arrived at through communication with a potential supervisor. The relative emphasis on methodology/theory vs. application may vary.

Suggested PhD research areas include:

  • Automating statistical advice
  • Change detection for complex data
  • Dimension reduction and factor modelling
  • Estimation of stochastic volatility models
  • Financial econometrics
  • Functional data analysis including functional time series
  • High-dimensional time series analysis
  • High-dimensional variable selection
  • Infectious disease modelling
  • Inference for sequential data including change detection in multiple data streams
  • Network time series analysis
  • Nonparametric and semiparametric regression
  • Non-stationary time series analysis
  • Reinforcement learning for time-dependent data
  • Robust statistical analysis for high-dimensional data
  • Shape-constrained methods
  • Spatial econometrics modelling
  • Spatio-temporal modelling
  • Statistical analysis of high-dimensional multi-type recurrent events

For more information, please see the Time Series and Statistical Learning research group