Title :
An all-MOS analog feedforward neural circuit with learning
Author :
Salam, Fathi M A ; Choi, Myung-Ryul
Author_Institution :
Dept. of Electr. Eng., Michigan State Univ., E. Lansing, MI, USA
Abstract :
An all-MOS circuit realization for a feedforward artificial neural network is described. An all-MOS realization of a modified learning rule is introduced. In addition to analytical verification the modified learning rule is shown, via computer code as well as SPICE simulations, to successfully store into the network any given analog values (within the permissible range). An all-MOS architecture for a prototype two-layer artificial neural network is specifically tested via SPICE simulations. The results demonstrate the learning capability of the all-MOS circuit realization and establish a VLSI modular architecture for composing a large-scale neural network system
Keywords :
MOS integrated circuits; VLSI; circuit analysis computing; digital simulation; linear integrated circuits; neural nets; SPICE simulations; VLSI modular architecture; all-MOS architecture; all-MOS circuit realization; analog feedforward neural circuit with learning; analytical verification; feedforward artificial neural network; large-scale neural network system; learning capability; modified learning rule; two-layer artificial neural network; Analog computers; Analytical models; Artificial neural networks; Circuit simulation; Computational modeling; Computer architecture; Computer networks; Computer simulation; SPICE; Virtual prototyping;
Conference_Titel :
Circuits and Systems, 1990., IEEE International Symposium on
Conference_Location :
New Orleans, LA
DOI :
10.1109/ISCAS.1990.112520
