Robust “Topological” Codes by Keeping Control of Internal Redundancy
Michael Haft · Physical Review Letters · 1998
Processing information under noisy conditions requires finding a tradeoff between coding a variety of different information and robust redundant coding of important information. We illustrate this information-theoretic requirement by some simple considerations. Following this, we set up information-theoretic plausible learning rules for a self-organizing network. Thereby, internal redundancy is controlled via anti-Hebbian learning based on an internal topology with a given correlation function. The emergence of maplike representation of sensory information is shown to be the consequence.