Title :
A parallel genetic/neural network learning algorithm for MIMD shared memory machines
Author :
Hung, S.L. ; Adeli, H.
Author_Institution :
Dept. of Civil Eng., Nat. Chiao Tung Univ., Hsinchu, Taiwan
fDate :
11/1/1994 12:00:00 AM
Abstract :
A new algorithm is presented for training of multilayer feedforward neural networks by integrating a genetic algorithm with an adaptive conjugate gradient neural network learning algorithm. The parallel hybrid learning algorithm has been implemented in C on an MIMD shared memory machine (Cray Y-MP8/864 supercomputer). It has been applied to two different domains, engineering design and image recognition. The performance of the algorithm has been evaluated by applying it to three examples. The superior convergence property of the parallel hybrid neural network learning algorithm presented in this paper is demonstrated
Keywords :
convergence of numerical methods; feedforward neural nets; genetic algorithms; image recognition; learning (artificial intelligence); parallel algorithms; parallel machines; Cray Y-MP8/864 supercomputer; MIMD shared memory machines; adaptive conjugate gradient neural network; convergence; engineering design; genetic algorithm; image recognition; learning algorithm; multilayer feedforward neural networks; parallel genetic/neural network; parallel hybrid learning algorithm; Backpropagation algorithms; Biological neural networks; Convergence; Design engineering; Feedforward neural networks; Genetic algorithms; Image recognition; Machine learning; Neural networks; Supercomputers;
Journal_Title :
Neural Networks, IEEE Transactions on
