학술논문
Modular neural network architectures for classification
Document Type
Conference
Author
Source
Proceedings of International Conference on Neural Networks (ICNN'96) Neural networks Neural Networks, 1996., IEEE International Conference on. 2:1279-1284 vol.2 1996
Subject
Language
Abstract
One of the major drawbacks of the current neural network generation is the inability to cope with the increase of size/complexity of classification tasks. Modular neural network classifiers attempt to solve this problem through a "divide and conquer" approach. However. The performance of the modular neural network classifiers is sensitive to efficiency of the "task decomposition" technique and the "multi-module decision-making" strategy. After a brief review of previous work with emphasis on five published modular classifiers-decoupled nets, ART-BP, hierarchical network, multiple experts, and multiple identical networks (majority vote and average output decisions)-this paper introduces the cooperative modular neural network (CMNN). The CMNN classifier outperforms the surveyed nets due to its novel task decomposition and multi-module decision-making techniques.