Title :
Design and sensitivity analysis of feed-forward neural ADC´s
Author :
Movahedian, Hamid ; Bakhtiar, Mehrdad Sharif
Author_Institution :
Dept. of Electr. Eng., Sharif Univ. of Technol., Tehran, Iran
Abstract :
In this paper, a one-layer and a two-layer Gray-coded feed-forward neural ADC with soft-limiting activation functions are introduced. The use of soft-limiting activation functions results in reduction of the total number of neurons, but at the same time it increases the sensitivity of various errors. The required number of bits for training the network derived from sensitivity analysis. A design example is also presented.
Keywords :
Gray codes; analogue-digital conversion; errors; feedforward neural nets; sensitivity analysis; Gray-coded feedforward neural ADC; errors sensitivity; neurons number reduction; one-layer feedforward neural ADC; sensitivity analysis; soft-limiting activation functions; two-layer feedforward neural ADC; Equations; Feedforward systems; Neural networks; Neurons; Sensitivity analysis; Stability; Switches;
Conference_Titel :
Circuits and Systems, 2004. ISCAS '04. Proceedings of the 2004 International Symposium on
Print_ISBN :
0-7803-8251-X
DOI :
10.1109/ISCAS.2004.1329933
