Document Server@UHasselt >
Research >
Research publications >

Please use this identifier to cite or link to this item: http://hdl.handle.net/1942/20662

Title: Positive Neural Networks in Discrete Time Implement Monotone-Regular Behaviors
Authors: Ameloot, Tom J.
Van den Bussche, Jan
Issue Date: 2015
Publisher: MIT PRESS
Citation: NEURAL COMPUTATION, 27 (12), p. 2623-2660
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 the absence of noise, the class of positive neural networks captures the so-called monotone-regular behaviors, which 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 cannot be implemented with zero delay.
Notes: [Ameloot, Tom J.] Hasselt Univ, B-3500 Hasselt, Belgium. Transnat Univ Limburg, B-3500 Hasselt, Belgium.
URI: http://hdl.handle.net/1942/20662
DOI: 10.1162/NECO_a_00789
ISI #: 000365674700005
ISSN: 0899-7667
Category: A1
Type: Journal Contribution
Validation: ecoom, 2016
Appears in Collections: Research publications

Files in This Item:

Description SizeFormat
preprint563.16 kBAdobe PDF

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.