Parametrizing analog multi-compartment neurons with genetic algorithms [version 2; peer review: 1 approved, 2 approved with reservations]

Background Finding appropriate model parameters for multi-compartmental neuron models can be challenging. Parameters such as the leak and axial conductance are not always directly derivable from neuron observations but are crucial for replicating desired observations. The objective of this study is...

Full description

Saved in:
Bibliographic Details
Main Authors: Raphael Stock, Eric Müller, Jakob Kaiser, Sebastian Schmitt, Johannes Schemmel
Format: Article
Language:English
Published: F1000 Research Ltd 2024-11-01
Series:Open Research Europe
Subjects:
Online Access:https://open-research-europe.ec.europa.eu/articles/3-144/v2
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1846127373873840128
author Raphael Stock
Eric Müller
Jakob Kaiser
Sebastian Schmitt
Johannes Schemmel
author_facet Raphael Stock
Eric Müller
Jakob Kaiser
Sebastian Schmitt
Johannes Schemmel
author_sort Raphael Stock
collection DOAJ
description Background Finding appropriate model parameters for multi-compartmental neuron models can be challenging. Parameters such as the leak and axial conductance are not always directly derivable from neuron observations but are crucial for replicating desired observations. The objective of this study is to replicate the attenuation behavior of an excitatory postsynaptic potential (EPSP) traveling along a linear chain of compartments on the analog BrainScaleS-2 neuromorphic hardware platform. Methods In the present publication we use genetic algorithms to find suitable model parameters. They promise parameterization without domain knowledge of the neuromorphic substrate or underlying neuron model. To validate the results of the genetic algorithms, a comprehensive grid search was conducted. Furthermore, trial-to-trial variations in the analog system are counteracted utilizing spike-triggered averaging. Results and conclusions The algorithm successfully replicated the desired EPSP attenuation behavior in both single and multi-objective searches illustrating the applicability of genetic algorithms to parameterize analog neuromorphic hardware.
format Article
id doaj-art-47696dff20554c2bafd84de4d8a1fcea
institution Kabale University
issn 2732-5121
language English
publishDate 2024-11-01
publisher F1000 Research Ltd
record_format Article
series Open Research Europe
spelling doaj-art-47696dff20554c2bafd84de4d8a1fcea2024-12-12T01:00:00ZengF1000 Research LtdOpen Research Europe2732-51212024-11-01320368Parametrizing analog multi-compartment neurons with genetic algorithms [version 2; peer review: 1 approved, 2 approved with reservations]Raphael Stock0https://orcid.org/0009-0008-5531-1072Eric Müller1Jakob Kaiser2https://orcid.org/0000-0002-3586-2634Sebastian Schmitt3Johannes Schemmel4Kirchhoff Institute for Physics, Heidelberg University, Heidelberg, 69120, GermanyEuropean Institute for Neuromorphic Computing, Heidelberg University, Heidelberg, 69120, GermanyKirchhoff Institute for Physics, Heidelberg University, Heidelberg, 69120, GermanyDepartment for Neuro- and Sensory Physiology, University Medical Center Göttingen, Göttingen, 37077, GermanyKirchhoff Institute for Physics, Heidelberg University, Heidelberg, 69120, GermanyBackground Finding appropriate model parameters for multi-compartmental neuron models can be challenging. Parameters such as the leak and axial conductance are not always directly derivable from neuron observations but are crucial for replicating desired observations. The objective of this study is to replicate the attenuation behavior of an excitatory postsynaptic potential (EPSP) traveling along a linear chain of compartments on the analog BrainScaleS-2 neuromorphic hardware platform. Methods In the present publication we use genetic algorithms to find suitable model parameters. They promise parameterization without domain knowledge of the neuromorphic substrate or underlying neuron model. To validate the results of the genetic algorithms, a comprehensive grid search was conducted. Furthermore, trial-to-trial variations in the analog system are counteracted utilizing spike-triggered averaging. Results and conclusions The algorithm successfully replicated the desired EPSP attenuation behavior in both single and multi-objective searches illustrating the applicability of genetic algorithms to parameterize analog neuromorphic hardware.https://open-research-europe.ec.europa.eu/articles/3-144/v2analog computing neuromorphic genetic algorithm multi-compartmenteng
spellingShingle Raphael Stock
Eric Müller
Jakob Kaiser
Sebastian Schmitt
Johannes Schemmel
Parametrizing analog multi-compartment neurons with genetic algorithms [version 2; peer review: 1 approved, 2 approved with reservations]
Open Research Europe
analog computing
neuromorphic
genetic algorithm
multi-compartment
eng
title Parametrizing analog multi-compartment neurons with genetic algorithms [version 2; peer review: 1 approved, 2 approved with reservations]
title_full Parametrizing analog multi-compartment neurons with genetic algorithms [version 2; peer review: 1 approved, 2 approved with reservations]
title_fullStr Parametrizing analog multi-compartment neurons with genetic algorithms [version 2; peer review: 1 approved, 2 approved with reservations]
title_full_unstemmed Parametrizing analog multi-compartment neurons with genetic algorithms [version 2; peer review: 1 approved, 2 approved with reservations]
title_short Parametrizing analog multi-compartment neurons with genetic algorithms [version 2; peer review: 1 approved, 2 approved with reservations]
title_sort parametrizing analog multi compartment neurons with genetic algorithms version 2 peer review 1 approved 2 approved with reservations
topic analog computing
neuromorphic
genetic algorithm
multi-compartment
eng
url https://open-research-europe.ec.europa.eu/articles/3-144/v2
work_keys_str_mv AT raphaelstock parametrizinganalogmulticompartmentneuronswithgeneticalgorithmsversion2peerreview1approved2approvedwithreservations
AT ericmuller parametrizinganalogmulticompartmentneuronswithgeneticalgorithmsversion2peerreview1approved2approvedwithreservations
AT jakobkaiser parametrizinganalogmulticompartmentneuronswithgeneticalgorithmsversion2peerreview1approved2approvedwithreservations
AT sebastianschmitt parametrizinganalogmulticompartmentneuronswithgeneticalgorithmsversion2peerreview1approved2approvedwithreservations
AT johannesschemmel parametrizinganalogmulticompartmentneuronswithgeneticalgorithmsversion2peerreview1approved2approvedwithreservations