DocumentCode
288576
Title
A simulation and training technique for analog neural network implementations
Author
Mundie, David B. ; Massengill, Lloyd W.
Author_Institution
Dept. of Electr. Eng., Memphis State Univ., TN, USA
Volume
3
fYear
1994
fDate
27 Jun-2 Jul 1994
Firstpage
1975
Abstract
A neural network simulation technique tailored for analog hardware implementation is discussed. Lookup tables are used to represent the nonideal properties of the analog circuitry. A weight perturbation training algorithm, well suited for hardware implementations, is presented. Simulated annealing schemes are shown to improve training performance associated with analog neural networks
Keywords
analogue processing circuits; circuit analysis computing; learning (artificial intelligence); neural nets; table lookup; analog circuitry; analog neural network; lookup tables; neural network simulation; simulated annealing; training; weight perturbation training algorithm; Analog circuits; Artificial neural networks; Circuit simulation; Computational modeling; Dynamic range; Neural network hardware; Neural networks; Pulse modulation; SPICE; Table lookup;
fLanguage
English
Publisher
ieee
Conference_Titel
Neural Networks, 1994. IEEE World Congress on Computational Intelligence., 1994 IEEE International Conference on
Conference_Location
Orlando, FL
Print_ISBN
0-7803-1901-X
Type
conf
DOI
10.1109/ICNN.1994.374464
Filename
374464
Link To Document