Design of experiments in regression models with correlated observations
This research project is in competition for funding with one or more projects available across the EPSRC Doctoral Training Partnership (DTP). Usually the projects which receive the best applicants will be awarded the funding. Find out more information about the DTP and how to apply.
Application deadline: 15 March 2019
Start date: 1 October 2019
The main aim of the project is to develop the theoretical and methodological principles for the problem of optimal experimental design in the case of correlated observations.
The design of experiments is an area of mathematical sciences which has direct economic impact. Optimal designs allow saving time, materials and other resources. New design methodologies are required due to growing complexity of real world problems, for example, technological changes in quality and quantity of measured data. The theory of the design of experiments involves statistical, optimization and simulation principles, various mathematical techniques and extensive computing.
The theory of statistical experimental design is an important link between the experimental and the modelling worlds. Designing experiments aims at extracting the maximal information from an experiment. Up to now the theory of experimental design mostly deals with uncorrelated observations. In most modern applications, however, the experimental observations are correlated. The whole philosophy of experimental design in this case becomes more complicated.
Project aims and methods
The project is interdisciplinary in nature and will include elements of theoretical probability and applied statistics. The main objective of the project is to develop the theoretical and methodological principles for the problem of optimal experimental design in the case of correlated observations.
Specifically, statistical procedures and computer software will be developed for solving specific classes of applied problems including optimal designs for spatial models, sequential designs for computer experiments, efficient designs for the evaluation of the pharmacokinetics and the pharmacodynamics of drugs.