Title :
FPGA implementation of bidirectional associative memory via simultaneous perturbation rule
Author :
Wakamura, Masatoshi ; Maeda, Yutaka
Author_Institution :
Dept. of Electr. Eng., Kansai Univ., Suita, Japan
Abstract :
Bidirectional associative memory (BAM) is a typical example of recurrent neural networks. Software implementation of BAM does not obtain sufficient speed in the operation. Therefore, hardware implementation, especially, FPGA (field programmable gate array) implementation of BAM is very promising. Originally, the weights of BAM are calculated by the patterns to be memorized. However, we adopt a recursive learning method via the simultaneous perturbation learning rule.
Keywords :
content-addressable storage; field programmable gate arrays; hardware-software codesign; learning (artificial intelligence); neural nets; FPGA; VHDL; bidirectional associative memory; recurrent neural networks; recursive learning; simultaneous perturbation learning rule; Associative memory; Circuit synthesis; Digital circuits; Field programmable gate arrays; Hardware design languages; Magnesium compounds; Neural networks; Neurons; Reverberation; Very high speed integrated circuits;
Conference_Titel :
SICE 2002. Proceedings of the 41st SICE Annual Conference
Print_ISBN :
0-7803-7631-5
DOI :
10.1109/SICE.2002.1196557
