Neural network utilizing logarithmic function and method of using same
作者： Shay-Ping Thomas Wang
DOI:
关键词: Time delay neural network 、 Logarithm 、 Artificial neural network 、 Echo state network 、 Artificial intelligence 、 Adder 、 Function (mathematics) 、 Stochastic neural network 、 Computer science 、 Probabilistic neural network 、 Algorithm
摘要: A neural network, which may be implemented either in hardware or software, is constructed of neurons neuron circuits each having only one significant processing element the form an adder. Each network further includes for applying a logarithmic function to its inputs and inverse-logarithmic outputs neurons. The utilizes training algorithm does not require repetitive yields global minimum given set input vectors.
lens.org 参考文章(39)
Daniel Hampel, Richard William Blasco, Multionomial processor system ,(1974)
Terry A. Jeeves, John H. Murphy, Digital neural network processing elements ,(1989)
Martin C. Peckerar, Christie R. K. Marrian, Cross entropy deconvolver circuit adaptable to changing convolution functions ,(1990)
Martin C. Peckerar, Christie R. K. Marrian, Maximum entropy deconvolver circuit based on neural net principles ,(1988)
Edwin W Herron, Robert D Hunter, David E Keefer, Multiplication apparatus in a data processing system with a variable length multiplier ,(1965)
Jean Rabasse, Series type adder for adding plural binary numbers ,(1974)
Ingwersen L, General purpose matrix processor with convolution capabilities ,(1970)
Albert A. Petrick, David E. Ratti, Paul M. Basehore, Neural processor apparatus ,(1992)
Thomas B. Brightman, Warren Ferguson, Method and apparatus for performing mathematical functions using polynomial approximation and a rectangular aspect ratio multiplier ,(1990)