Title :
An analog neural network circuit with simultaneous perturbation learning rule
Author :
Maeda, Yutaka ; Kanata, Yakichi
Author_Institution :
Dept. of Electr. Eng., Kansai Univ., Osaka, Japan
Abstract :
In this paper, we describe an implementation of an analog neural network system using the simultaneous perturbation learning rule. Compared with usage of the ordinary back-propagation method, we could easily implement the simultaneous perturbation learning rule. The exclusive OR problem and a simple function problem are shown
Keywords :
analogue processing circuits; learning (artificial intelligence); neural nets; perturbation techniques; analog neural network circuit; exclusive OR problem; simultaneous perturbation learning rule; sin function problem; Circuits; Education; Electronic mail; Feedforward neural networks; Gradient methods; Neural network hardware; Neural networks; Newton method;
Conference_Titel :
Circuits and Systems, 1998. ISCAS '98. Proceedings of the 1998 IEEE International Symposium on
Conference_Location :
Monterey, CA
Print_ISBN :
0-7803-4455-3
DOI :
10.1109/ISCAS.1998.703911
