Autor:	Takahashi, Yoshikane
Předmět:	ARTIFICIAL neural networks RATE distortion theory INFORMATION theory MATHEMATICS
Zdroj:	IEEE Transactions on Circuits & Systems Part I: Fundamental Theory & Applications; Feb2000, Vol. 47 Issue 2, p166, 19p, 1 Black and White Photograph, 6 Diagrams
Abstrakt:	Deals with a study which developed a mathematical solution to the network construction problems (NCP) in neural networks using rate distortion theory from information theory and linear algebra. Formulating the NCP; Constructing a minimal classifier; Constructing a minimal frame; Solving the NCP; Procedure as a solution to the NCP and network training problem; Conclusion.
Databáze:	Complementary Index
Externí odkaz:	Zobrazit plný text záznamu