Additive Composition of Supervised Self Organizing Maps

The learning of complex relationships can be decomposed into several neural networks. The modular organization is determined by prior knowledge of the problem that permits to splitthe processing into tasks ofsmall dimensionality.The sub-tasks can be implemented with neural networks, although the learning examples cannot be used anymore to supervise directly each ofthe networks.This article addresses the problem of learning in a modular context, developing inparticularadditivecompositions. Asimple ruleallows de

• Type: Journal
• Year: 2002
• Where: NPL
• Authors: Jean-Luc Buessler, Jean-Philippe Urban, Julien Gresser
• Keywords: Complex Relationships, Modular Organization, Neural Networks, NPL 2002