Stability of discrete time recurrent neural networks and nonlinear optimization problems

DOI10.1016/J.NEUNET.2015.10.013zbMATH Open1395.39008arXiv1503.01818OpenAlexW1631735240WikidataQ50758046 ScholiaQ50758046MaRDI QIDQ1667540FDOQ1667540

Jayant Singh, Nikita E. Barabanov

Publication date: 30 August 2018

Published in: Neural Networks (Search for Journal in Brave)

Abstract: We consider the method of Reduction of Dissipativity Domain to prove global Lyapunov stability of Discrete Time Recurrent Neural Networks. The standard and advanced criteria for Absolute Stability of these essentially nonlinear systems produce rather weak results. The method mentioned above is proved to be more powerful. It involves a multi-step procedure with maximization of special nonconvex functions over polytopes on every step. We derive conditions which guarantee an existence of at most one point of local maximum for such functions over every hyperplane. This nontrivial result is valid for wide range of neuron transfer functions.

Full work available at URL: https://arxiv.org/abs/1503.01818

zbMATH Keywords

optimizationLyapunov stabilityrecurrent neural network

Mathematics Subject Classification ID

Stability theory for difference equations (39A30) Applications of difference equations (39A60)

Cites Work

• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• New Results on Stability of Discrete-Time Systems With Time-Varying State Delay
• Image-processing algorithms realized by discrete-time cellular neural networks and their circuit implementations
• The stabilizing solution of the discrete algebraic riccati equation
• Hermitian solutions of the discrete algebraic Riccati equation
• On critical cases in the theory of absolute stability of controlled systems
• A new method for stability analysis of nonlinear discrete-time systems

Cited In (11)

• Title not available (Why is that?)
• Title not available (Why is that?)
• Four discrete-time ZD algorithms for dynamic linear matrix-vector inequality solving
• Local stability conditions for discrete-time cascade locally recurrent neural networks
• On the stability properties of gated recurrent units neural networks
• A critical case for the spiral stability for \(2\times 2\) discontinuous systems and an application to recursive neural networks
• On the S-instability and degeneracy of discrete deep learning models
• Dynamical stability conditions for recurrent neural networks with unsaturating piecewise linear transfer functions
• Lur'e systems with multilayer perceptron and recurrent neural networks: absolute stability and dissipativity
• Delay-independent stability criteria for complex-valued BAM neutral-type neural networks with time delays
• Design and analysis of two discrete-time ZD algorithms for time-varying nonlinear minimization

This page was built for publication: Stability of discrete time recurrent neural networks and nonlinear optimization problems

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1667540)