Brain Modelling and Applications

Work in this group spans empirical psychological research, computer modelling of brains and mind and applications of computer models of brain processing.

Rainer Spiegel's work on sequence learning is paradigmatic here. The research began in the Cambridge University Psychology Department (who still collaborate on the project) and concerned experiments exploring humans' capacity for this kind of learning. To understand the possible mechanisms behind the empirical results, Rainer designs and implements computer systems and compares the behaviour of the artificial systems and  the experimental results. This work has generated interest from Psychologists and Computer Scientists.

Mark Bishop has worked towards a new conceptual framework for Neural Networks, Stochastic Diffusion Processes, SDPs. and has received significant funding to apply traditional Neural technologies to a variety of real industrial problems - most significantly in the field of Colour Physics and Mobile Robot Localisation.

Nikolay Nikolaev works on Polynomial neural networks (PNN), which are multilayer perceptrons of neuron-like units which produce high-order multivariate polynomial mappings. These are tree-structured hierarchical cascades of first-order and second-order activation polynomials in the nodes, and input variables passed from the leaves. He has also worked on medical and financial applications.