Talk on supervised data clustering

Prof. Sossa from Mexico’s National Polytechnic Institute is visiting Jena and presents recent work on supervised data clustering and its application to training of neural networks.

Abstract:

In this talk I will describe a new supervised data clustering method and how it can be applied into for training of neural networks. I particular, I describe how the proposal can be used to train Lattice Neural Networks with Dendritic Processing (Morphological Neural Networks) and Radial Basis Function Neural Networks. In the second case, I introduce a new architecture. In both cases, I provide examples where the trained neural networks present a comparable and, in some case, a superior performance than similar reported approaches reported in literature.