DocumentCode :
3565860
Title :
A general-purpose neural network with on-chip BP learning
Author :
Lu, Chun ; Shi, Bingxue ; Chen, Lu
Author_Institution :
Inst. of Microelectron., Tsinghua Univ., Beijing, China
Volume :
2
fYear :
2002
fDate :
6/24/1905 12:00:00 AM
Abstract :
A general purpose expandable neural network chip with on-chip BP (back-propagation) learning is designed. The unit chip has 4 neurons and 16 synapses. A large-scale neural network with arbitrary layers and discretional neurons per layer can be constructed by combining many unit chips. A novel neuron circuit with programmable parameters is proposed. It generates not only the sigmoid function but also its derivative. The neuron has a push-pull output stage to gain strong driving ability in both charge and discharge processes, which is very important in heavy load situations. The unit chip is fabricated with a standard 0.5-μm CMOS, double-poly, double-metal technology. The learning system itself can be used as a refresh tool to keep the weight value right. The results of parity experiments show that it can accomplish on-chip BP learning.
Keywords :
CMOS analogue integrated circuits; backpropagation; integrated circuit design; integrated circuit testing; neural chips; programmable circuits; 0.5 micron; CMOS double-poly double-metal unit chip technology; arbitrary layers; charge processes; discharge processes; discretional neurons per layer; general purpose expandable neural network chip; heavy load situations; large-scale neural network; learning system; neuron circuit; neuron driving ability; neurons; on-chip BP learning; on-chip back-propagation learning; parity experiments; programmable parameters; push-pull output stage; refresh tool; sigmoid function; sigmoid function derivative; synapses; weight value; CMOS technology; Circuits; Fabrication; Large-scale systems; Microelectronics; Network-on-a-chip; Neural networks; Neurons; Signal generators; System-on-a-chip;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Circuits and Systems, 2002. ISCAS 2002. IEEE International Symposium on
Print_ISBN :
0-7803-7448-7
Type :
conf
DOI :
10.1109/ISCAS.2002.1011039
Filename :
1011039
Link To Document :
بازگشت