DocumentCode :
2204715
Title :
Discussion on the methodology of neural network hardware design and implementation
Author :
Wang, Xingang ; Ma, Zuocheng ; Lu, Huaxiang ; Wang, Shoujue
Author_Institution :
Inst. of Intelligent Inf. Syst., Zhejiang Univ. of Technol., Hangzhou, China
Volume :
1
fYear :
2001
fDate :
22-25 Oct. 2001
Firstpage :
113
Abstract :
As a suitable calculating method for parallel computation, neural networks have been widely used in applications. It has been pointed out that the main problem of neural network hardware implementation lies in three aspects. To solve these problems, the way of hardware implementation of neural networks has been discussed. A novel basic mathematical model of the neuron has been introduced. Using this new model of the neuron, the flexibility of the implemented neural network hardware will be greatly improved.
Keywords :
integrated circuit design; neural nets; methodology; neural network hardware design; neural network hardware implementation; Application software; Biological neural networks; Computer networks; Information systems; Intelligent networks; Intelligent systems; Mathematical model; Neural network hardware; Neural networks; Neurons;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Solid-State and Integrated-Circuit Technology, 2001. Proceedings. 6th International Conference on
Conference_Location :
Shanghai, China
Print_ISBN :
0-7803-6520-8
Type :
conf
DOI :
10.1109/ICSICT.2001.981435
Filename :
981435
Link To Document :
بازگشت