The handbook of brain theory and neural networks
An Introduction to Morphological Neural Networks
ICPR '96 Proceedings of the International Conference on Pattern Recognition (ICPR '96) Volume IV-Volume 7472 - Volume 7472
| Hi-index | 0.00 |
In this paper, a novel model of Morphological Neural Networks (MNN) is presented. MNNs use neurons with dendritic structure that bear a more faithful resemblance to the biological neurons of the celebral cortex than those used in various artificial neural networks. MNNs have the high capability of resolving some difficult non-linear problems and are getting more and more applicable. In this work we develop a new model of MNNs that are more powerful since each dendrite can work on a different orthonormal basis than the other dendrites in order to optimize the performance of the Orthonormal Basis Morphological Neural Network (OBMNN). OBMNNs compress automatically their size and show significantly better learning capabilities than the regular MNNs, which are a sub-group of OBMNNs. Validation experimental results are presented that demonstrate superior learning performance of OBMNNs over regular MNNs.