Professor Peter Challenor

Peter Challenor has broad interests; mainly about uncertainty in the world. These range from the statistical analysis of complex numerical models (such as those used to simulate climate or in healthcare and engineering) to the interpolation of noisy data.

To carry out uncertainty quantification on complex numerical models (mainly ODEs and PDEs) that are too expensive for Monte Carlo methods we build emulators. These are stochastic surrogates models fitted to a carefully designed set ofmodel runs.our emulators are usually Gaussian processes. The resulting Gaussian process is orders of magnitude faster than the original model. It can then be used for prediction (including exploring extreme cases), sensitivity analysis, uncertainty analysis and model calibration.

Peter is the principal investigator on the EPSRC Excalbur poject EXA-UQ. This is concerned with uncertainty qunatification on the next generation of High Peformance Computers. Uncertainty quantification uses a carefully designed experimentHe is co-investigator on the EPSRC Hub for Quantitative Modelling in Healthcare and the NERC funded RENEW project about the interaction of people and biodiversity.