DocumentCode :
298565
Title :
The impact of VLSI fabrication on neural learning
Author :
Card, H.C. ; McNeill, D.K. ; Schneider, C.R. ; Schneider, R.S.
Author_Institution :
Dept. of Electr. & Comput. Eng., Manitoba Univ., Winnipeg, Man., Canada
Volume :
2
fYear :
1995
fDate :
30 Apr-3 May 1995
Firstpage :
985
Abstract :
The fabrication of silicon versions of artificial neural learning algorithms in existing VLSI processes introduces a variety of concerns which do not exist in a theoretical system. These include such well known circuit properties as noise, variations and nonlinearity of fabricated devices, arithmetic inaccuracy, and capacitive decay. The supervised learning algorithm-contrastive Hebbian learning, and unsupervised soft competitive learning have demonstrated their resiliency in the presence of these effects as observed in 1.2 μm CMOS circuits employing Gilbert multipliers. It has been found that the learning circuits will operate correctly in the presence of offset errors in analog multipliers and adders, if thresholding is applied when performing weight updates
Keywords :
CMOS analogue integrated circuits; Hebbian learning; VLSI; analogue multipliers; neural chips; unsupervised learning; 1.2 micron; CMOS circuits; Gilbert multipliers; Si; VLSI fabrication; adders; analog multipliers; arithmetic inaccuracy; capacitive decay; contrastive Hebbian learning; neural learning algorithm; noise; nonlinearity; offset errors; silicon; supervised learning; thresholding; unsupervised soft competitive learning; Arithmetic; Artificial neural networks; CMOS analog integrated circuits; CMOS technology; Circuit noise; Fabrication; Hebbian theory; Neurons; Silicon; Very large scale integration;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Circuits and Systems, 1995. ISCAS '95., 1995 IEEE International Symposium on
Conference_Location :
Seattle, WA
Print_ISBN :
0-7803-2570-2
Type :
conf
DOI :
10.1109/ISCAS.1995.519931
Filename :
519931
Link To Document :
بازگشت