MTiling A Constructive Neural Network Learning Algorithm for Multi Category Pattern Classi cation

Constructive learning algorithms o er an approach for incremental construction of potentially near minimal neural network architectures for pattern classi cation tasks Such algorithms help overcome the need for ad hoc and often inappropriate choice of network topology in the use of algorithms that search for a suitable weight setting in an otherwise a priori xed network architecture Several such algorithms proposed in the literature have been shown to converge to zero classi cation errors under certain as sumptions on a nite non contradictory training set in a category classi cation problem This paper presents MTiling a multi category extension of Tiling algorithm M ezard Nadal We establish the convergence of MTiling to zero classi cation error on a multi category pattern classi cation task Results of experiments with non linearly separable multi category data sets demonstrate the feasibility of this approach to multi category pattern classi cation and also suggest several interesting directions for future research