Positive neural networks in discrete time implement monotone-regular behaviors

DOI10.1162/NECO_A_00789zbMATH Open1414.92004DBLPjournals/neco/AmelootB15arXiv1502.06094WikidataQ86546317 ScholiaQ86546317MaRDI QIDQ5380357FDOQ5380357

Authors: Tom J. Ameloot, Jan Van den Bussche 

Publication date: 4 June 2019

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

Abstract: We study the expressive power of positive neural networks. The model uses positive connection weights and multiple input neurons. Different behaviors can be expressed by varying the connection weights. We show that in discrete time, and in absence of noise, the class of positive neural networks captures the so-called monotone-regular behaviors, that are based on regular languages. A finer picture emerges if one takes into account the delay by which a monotone-regular behavior is implemented. Each monotone-regular behavior can be implemented by a positive neural network with a delay of one time unit. Some monotone-regular behaviors can be implemented with zero delay. And, interestingly, some simple monotone-regular behaviors can not be implemented with zero delay.

Full work available at URL: https://arxiv.org/abs/1502.06094

Recommendations

• Positive automata networks
• Artificial Intelligence and Soft Computing - ICAISC 2004
• scientific article; zbMATH DE number 2242764
• Global dynamics of discrete neural networks allowing non-monotonic activation functions
• Symmetric discrete universal neural networks

zbMATH Keywords

discrete timeconnection weightspositive neural networksmonotone-regular behavior

Mathematics Subject Classification ID

Neural networks for/in biological studies, artificial life and related topics (92B20)

Cites Work

• Title not available (Why is that?)
• Title not available (Why is that?)
• The monotone circuit complexity of Boolean functions
• On the computational power of neural nets
• Monotone circuits for monotone weighted threshold functions
• General-Purpose Computation with Neural Networks: A Survey of Complexity Theoretic Results
• Title not available (Why is that?)
• Links between probabilistic automata and hidden Markov models: probability distributions, learning models and induction algorithms
• Title not available (Why is that?)
• Efficient simulation of finite automata by neural nets
• Lower Bounds for the Computational Power of Networks of Spiking Neurons
• Constructing deterministic finite-state automata in recurrent neural networks
• Optimal simulation of automata by neural nets
• Theory of neuromata
• On the Classification Capability of Sign-Constrained Perceptrons

Cited In (4)

• Positive solutions and exponential stability of nonlinear time-delay systems in the model of BAM-Cohen-Grossberg neural networks
• Robust exponential stabilization of positive uncertain switched neural networks with actuator saturation and sensor faults
• \(\ell_1\)-gain filter design of discrete-time positive neural networks with mixed delays
• Exponential attractivity of positive inertial neural networks in bidirectional associative memory model with heterogeneous delays

This page was built for publication: Positive neural networks in discrete time implement monotone-regular behaviors

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5380357)