Title :
Structural simplification of FMLP
Author_Institution :
Lab. for Automated Reasoning & Programming, Chengdu Inst. of Comput. Application, China
Abstract :
The number of fuzzy rules directly determines the complexity and efficiency of a fuzzy multilayer perceptron (FMLP). Based on the neural network self-configuring learning (NNSCL) algorithm, the NNSCL-I algorithm is obtained by using the generalized inverse matrix (GIM) algorithm to adjust the remaining weights after pruning neurons. The NNSCL-I algorithm is applied in the rule-reasoning layer of the FMLP to simplify its rules and structure with no degradation in the original performance. Experimental results show the effectiveness and the feasibility of the algorithm.
Keywords :
fuzzy neural nets; learning (artificial intelligence); matrix inversion; multilayer perceptrons; planning (artificial intelligence); self-organising feature maps; FMLP; GIM algorithm; NNSCL-I algorithm; complexity; fuzzy multilayer perceptron; fuzzy rules; generalized inverse matrix; neural network self-configuring learning; rule-reasoning layer; structural simplification; Computer applications; Dispersion; Feedforward neural networks; Fuzzy set theory; Laboratories; Multi-layer neural network; Multilayer perceptrons; Neural networks; Neurons; Statistics;
Conference_Titel :
Communications, Circuits and Systems, 2005. Proceedings. 2005 International Conference on
Print_ISBN :
0-7803-9015-6
DOI :
10.1109/ICCCAS.2005.1495265
