Место издания:ICSC Academic Press International Computer Science Conventions Canada/Switzerland
Первая страница:113
Последняя страница:116
Аннотация:The paper describes main principles of a novel algorithm for creation of hierarchical neural network classifiers. The algorithm is based on a modification of error backpropagation, combining supervised learning with self-organization. Recursive use of the algorithm results in creation of compact and computationally effective self-organized structures of neural classifiers. The algorithm performance has been tested on model data and on real-world problems.