Plasticity-Mediated Competitive Learning

Differentiation between the nodes of a competitive learning network is conventionally achieved through competition on the basis of neural activity. Simple inhibitory mechanisms are limited to sparse representations, while decorrelation and factorization schemes that support distributed representations are computationally unattractive. By letting neural plasticity mediate the competitive interaction instead, we obtain diffuse, nonadaptive alternatives for fully distributed representations. We use this technique to simplify and improve our binary information gain optimization algorithm for feature extraction (Schraudolph and Sejnowski, 1993); the same approach could be used to improve other learning algorithms.