Activation functions of generalized (L)-type in neural networks theory

Nasser Eddine Tatar*

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

In this work we enlarge the class of admissible activation functions, in a Hpfield neural network system, leading to exponential stability. Most of the existing treatments so far consider only Lipschitz continuous activation functions. Here, we discuss the case of functions of “generalized” (L)-type which are not necessarily Lipschitz continuous and systems with variable coefficients. The case of constant coefficients is given as a particular case.

Original languageEnglish
Pages (from-to)199-209
Number of pages11
JournalAdvanced Mathematical Models and Applications
Volume4
Issue number3
StatePublished - 2019

Bibliographical note

Publisher Copyright:
© 2019, Jomard Publishing. All rights reserved.

Keywords

  • (L)-type kernel
  • Activation function
  • Exponential stabilization
  • Neural network

ASJC Scopus subject areas

  • Statistics and Probability
  • Numerical Analysis
  • Modeling and Simulation
  • Control and Optimization
  • Computational Mathematics
  • Applied Mathematics

Fingerprint

Dive into the research topics of 'Activation functions of generalized (L)-type in neural networks theory'. Together they form a unique fingerprint.

Cite this