Complex domain backpropagation

Author

Georgiou, George M. ; Koutsougeras, Cris

Author_Institution

Dept. of Comput. Sci., Tulane Univ., New Orleans, LA, USA

Volume

39

Issue

5

fYear

1992

fDate

5/1/1992 12:00:00 AM

Firstpage

330

Lastpage

334

Abstract

The backpropagation algorithm is extended to complex domain backpropagation (CDBP) which can be used to train neural networks for which the inputs, weights, activation functions, and outputs are complex-valued. Previous derivations of CDBP were necessarily admitting activation functions that have singularities, which is highly undesirable. In the derivation, CDBP is derived so that that it accommodates classes of suitable activation functions. One such function is found and the circuit implementation of the corresponding neuron is given. CDBP hardware circuits can be used to process sinusoidal signals all at the same frequency (phasors)

Keywords

analogue computer circuits; computerised signal processing; learning systems; neural nets; activation functions; complex domain backpropagation; hardware circuits; neural network training; phasors; sinusoidal signals processing; Backpropagation algorithms; Circuits; Feedforward systems; Frequency; Least squares approximation; Neural network hardware; Neural networks; Neurons; Nonhomogeneous media; Signal processing;

fLanguage

English

Journal_Title

Circuits and Systems II: Analog and Digital Signal Processing, IEEE Transactions on

Publisher

ieee

ISSN

1057-7130

Type

jour

DOI

10.1109/82.142037

Filename

142037