Input Layer Optimization of Neural Networks by Sensitivity Analysis and Its Application to Recognition of Numerals.

Autor:	Takenaga, Hiroshi, Abe, Shigeo, Takato, Masao, Kayama, Masahiro, Kitamura, Tadaaki, Okuyama, Yoshiyuki
Předmět:	ARTIFICIAL neural networks ARTIFICIAL intelligence COMPUTERS SENSITIVITY theory (Mathematics) NUMERALS AUTOMATION COMPUTER systems INPUT-output analysis
Zdroj:	Electrical Engineering in Japan; 1991, Vol. 111 Issue 4, p130-138, 9p
Abstrakt:	Focuses on the input layer optimization of neural networks by sensitivity analysis and its application to recognition of numerals. Usage of image recognition equipment in the field of factory automation for electronic parts assembly and product defect tests; Development of a recognition algorithm for such application; Discussion of the input-to-output sensitivity of the neural net.
Databáze:	Complementary Index
Externí odkaz:	Zobrazit plný text záznamu Plný text