Neurocomputational models of hippocampus-dependent place learning and navigation
Humans and other animals can readily remember significant places and associated events and return to these places as appropriate. Experimentally, studies of the neuro-psychological mechanisms underlying place learning and navigation offer unique opportunities, because similar tests can be used in rodent models and human participants. Studies in rodent models have led to a detailed understanding of the neuro-psychological mechanisms of place memory, and the importance of the hippocampus for place learning and navigation in humans and other animals is well-established. This project aims to develop quantitative models describing how neurons in the hippocampus and associated brain areas give rise to place learning and navigation, and construct an in silico model for testing ideas about functional mechanisms. The project brings together behavioural neuroscience expertise on hippocampal function and place learning (T Bast, Psychology) with expertise in mathematical and computational neuroscience (S Coombes, Mathematical Sciences). A particular emphasis will be on the hippocampal learning-behaviour translation: how place information (as encoded, for example, by hippocampal place cells) is related to decision making processes and, ultimately, translated into motor behaviour (for example, by way of interactions with prefrontal and subcortical circuits). From a mathematical perspective the project will develop new neurocomputational models of hippocampus-dependent place learning and navigation using tools from stochastic optimal control, reinforcement learning theory, dynamical systems and computational neuroscience.
Eligibility requirements: 1st class Mathematics degree (or other highly mathematical field), preferably MMath/MSc, or equivalent (exceptionally, a 2:1 or equivalent may be accepted).
Funding: UK/EU students - Full Fees and Stipend at the RCUK rate (£14,296 p.a.)
Steve CoombesSchool of Mathematical Sciences, University of Nottingham