| Brackley Christopher |
| Email |
C.A.Brackley warwick.ac.uk |
| Gender |
Male |
| Nationality |
United Kingdom |
| Year of birth |
1982 |
| Department |
Department of Physics |
| University |
University of Warwick |
| Country (of work) |
United Kingdom |
| Simulation Project Title |
Learning in simple network models with inhomogeneous connection weights. |
| Type of model |
Abstract neural network model |
| Description |
There has been a recent renewed interest in networks of simple two state neuron like
objects with simple dynamics including refractoriness and noise (eg. PNAS 102 p9948). I
propose to introduce spacial structure with homogeneous local connections and random sparse
long range connections to such a model, and include a learning algorithm. I hope to be able to
qualitatively replicate behaviour seen in experiment such as peaks of synchronous activity and
cortical songs (eg. see SCIENCE 304 p559). |
| Why complementary |
My current research focuses on introducing inhomogeneity in neural connections in a mean
field firing rate model of neural tissue. This project uses the same idea of inhomogeneity in
connections, but on a smaller length scale. By using very simple dynamics it should be
possible to identify which features of the model give rise to particular aspects of the
system's behaviour, which may not be seen in a mean field approach. |
| Tutor |
Arvind Kumar |