Kernal Methods for Growth Domain Modelling and Experimental Design

Experimental research into the growth of microbial pathogens forms a vital component of efforts to ensure food safety, however statistical analysis of the results remains somewhat crude. We aim to develop a methodology for improved statistical analysis and modelling of experimental data. <P>
The work will benefit from recent advances in kernel methods, currently the one of the most active areas in the field of machine learning. The adoption of a Bayesian framework will support active learning, whereby a model of the available data is used to direct the design of further experimental work to obtain the greatest decrease in the uncertainty in predictions made by the subsequent generation of growth domain models.