Bio-Inspired Architectures Substantially Reduce the Memory Requirements of Neural Network Models

dc.contributor.authorDalgaty, Thomas
dc.contributor.authorMiller, John P.
dc.contributor.authorVianello, Elisa
dc.contributor.authorCasas, Jérôme
dc.date.accessioned2022-09-06T20:48:55Z
dc.date.available2022-09-06T20:48:55Z
dc.date.issued2021-02
dc.description.abstractWe propose a neural network model for the jumping escape response behavior observed in the cricket cercal sensory system. This sensory system processes low-intensity air currents in the animal's immediate environment generated by predators, competitors, and mates. Our model is inspired by decades of physiological and anatomical studies. We compare the performance of our model with a model derived through a universal approximation, or a generic deep learning, approach, and demonstrate that, to achieve the same performance, these models required between one and two orders of magnitude more parameters. Furthermore, since the architecture of the bio-inspired model is defined by a set of logical relations between neurons, we find that the model is open to interpretation and can be understood. This work demonstrates the potential of incorporating bio-inspired architectural motifs, which have evolved in animal nervous systems, into memory efficient neural network models.en_US
dc.identifier.citationDalgaty T, Miller JP, Vianello E and Casas J (2021) Bio-Inspired Architectures Substantially Reduce the Memory Requirements of Neural Network Models. Front. Neurosci. 15:612359.en_US
dc.identifier.issn1662-453X
dc.identifier.urihttps://scholarworks.montana.edu/handle/1/17078
dc.language.isoen_USen_US
dc.publisherFrontiers Media SAen_US
dc.rightscc-byen_US
dc.rights.urihttps://creativecommons.org/licenses/by/4.0/en_US
dc.subjectneural networken_US
dc.titleBio-Inspired Architectures Substantially Reduce the Memory Requirements of Neural Network Modelsen_US
dc.typeArticleen_US
mus.citation.extentfirstpage1en_US
mus.citation.extentlastpage17en_US
mus.citation.journaltitleFrontiers in Neuroscienceen_US
mus.citation.volume15en_US
mus.data.thumbpage4en_US
mus.identifier.doi10.3389/fnins.2021.612359en_US
mus.relation.collegeCollege of Letters & Scienceen_US
mus.relation.departmentMicrobiology & Cell Biology.en_US
mus.relation.universityMontana State University - Bozemanen_US

Files

Original bundle
Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
dalgaty-neural-2021.pdf
Size:
3.79 MB
Format:
Adobe Portable Document Format
Description:
License bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
license.txt
Size:
826 B
Format:
Item-specific license agreed upon to submission
Description:
Copyright (c) 2002-2022, LYRASIS. All rights reserved.