A theory of self-organising neural networks
MANNA '95 Proceedings of the first international conference on Mathematics of neural networks : models, algorithms and applications: models, algorithms and applications
Stochastic Complexity in Statistical Inquiry Theory
Stochastic Complexity in Statistical Inquiry Theory
Hi-index | 0.00 |
The aim of this paper is to show how the problem of optimising the processing chain in a classifier combiner network can be recast as a Markov chain optimisation problem. A summary of the application of coding theory to Markov chains and to encoder networks is presented, and the key idea of optimising the joint probability density of the state of the processing chain is stated. An example of the application of these ideas to processing data from multiple correlated sources (i.e. a hierarchically correlated phase screen) is then given.