Generalization characteristics of complex-valued feedforward neural networks in relation to signal coherence

Akira Hirose, Shotaro Yoshida

Research output: Contribution to journalArticlepeer-review

115 Citations (Scopus)

Abstract

Applications of complex-valued neural networks (CVNNs) have expanded widely in recent years-in particular in radar and coherent imaging systems. In general, the most important merit of neural networks lies in their generalization ability. This paper compares the generalization characteristics of complex-valued and real-valued feedforward neural networks in terms of the coherence of the signals to be dealt with. We assume a task of function approximation such as interpolation of temporal signals. Simulation and real-world experiments demonstrate that CVNNs with amplitude-phase-type activation function show smaller generalization error than real-valued networks, such as bivariate and dual-univariate real-valued neural networks. Based on the results, we discuss how the generalization characteristics are influenced by the coherence of the signals depending on the degree of freedom in the learning and on the circularity in neural dynamics.

Original languageEnglish
Article number6138313
Pages (from-to)541-551
Number of pages11
JournalIEEE Transactions on Neural Networks and Learning Systems
Volume23
Issue number4
DOIs
Publication statusPublished - 2012
Externally publishedYes

Keywords

  • Complex-valued neural network
  • function approximation
  • generalization
  • supervised learning

ASJC Scopus subject areas

  • Software
  • Computer Science Applications
  • Computer Networks and Communications
  • Artificial Intelligence

Fingerprint Dive into the research topics of 'Generalization characteristics of complex-valued feedforward neural networks in relation to signal coherence'. Together they form a unique fingerprint.

Cite this