DocumentCode :
2694607
Title :
Structural learning of neural networks by differential evolution with degeneration using mappings
Author :
Takahama, Tetsuyuki ; Sakai, Setsuko ; Hara, Akira ; Iwane, Noriyuki
Author_Institution :
Hiroshima City Univ., Hiroshima
fYear :
2007
fDate :
25-28 Sept. 2007
Firstpage :
3434
Lastpage :
3441
Abstract :
Structural learning, in which the structure of estimation systems are optimized, has been actively studied in researches on supervised learning of neural networks and fuzzy rules. The GAd (genetic algorithm with degeneration) is one of the structural learning methods, which is modeled on genetic damage and degeneration. In GAd, a gene is defined by a pair of a normal value and a damaged rate that shows how much the gene is damaged. Simple one-point crossover and Gaussian mutation are adopted to deal with the pair. It was very difficult to incorporate newly proposed genetic operations such as blend crossover in GA or operations in differential evolution (DE). In this study, we propose a new idea to incorporate such operations by unifying the values according to a mapping, applying operations and separating the values according to the inverse mapping. This idea is applied to differential evolution, which is known to be an efficient and robust algorithm and DEd (differential evolution with degeneration) is proposed. To show the advantage of DEd, it is applied to the structural learning of a simple test function and neural networks. It is shown that DEd is more robust to high degeneration pressure and can find better estimation models faster, which have less number of parameters and less estimation errors, than GAd.
Keywords :
Gaussian processes; fuzzy set theory; genetic algorithms; learning (artificial intelligence); neural nets; Gaussian mutation; differential evolution with degeneration; estimation systems; fuzzy rules; genetic algorithm with degeneration; genetic damage; inverse mapping; neural networks; structural learning; supervised learning; Estimation error; Fuzzy neural networks; Fuzzy systems; Genetic algorithms; Genetic mutations; Learning systems; Neural networks; Robustness; Supervised learning; Testing;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Evolutionary Computation, 2007. CEC 2007. IEEE Congress on
Conference_Location :
Singapore
Print_ISBN :
978-1-4244-1339-3
Electronic_ISBN :
978-1-4244-1340-9
Type :
conf
DOI :
10.1109/CEC.2007.4424916
Filename :
4424916
Link To Document :
بازگشت