Constructive Neural Network Learning Algorithms for Multi-Category Pattern Classification

Constructive learning algorithms offer an approach for incremental construction of potentially near-minimal neural network architectures for pattern classification 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 fixed network architecture. Several such algorithms proposed in the literature have been shown to converge to zero classification errors (under certain assumptions) on a finite, non-contradictory training set in a 2-category classification problem. This paper explores multi-category extensions of several constructive neural network learning algorithms for pattern classification. In each case, we establish the convergence to zero classification errors on a multi-category classification task (under certain assumptions). Results of experiments with non linearly separable multi-category data sets demonstrate the feasibility of this approach to multi-category pattern classification and also suggest several interesting directions for future research