Measuring ensemble consistency without measuring tuning curves

Jadin C. Jackson; A. David Redish

doi:10.1016/j.neucom.2004.01.028

Measuring ensemble consistency without measuring tuning curves

Jadin C. Jackson, A. David Redish

Neuroscience

Research output: Contribution to journal › Article › peer-review

Abstract

An important question in information processing is the extent to which neural firing patterns remain consistent while processing representations. Transient changes in representational consistency can provide clues to the dynamics of neural processing. We present a generalized framework for measuring the consistency of a neuronal representation that does not require explicit knowledge of the parameters encoded by the ensemble. It requires only neuronal ensembles and a training set of neuronal activity that samples behavioral parameters equally. This will be useful in structures where the behavioral parameters signalled by the neural activity are controversial or unknown.

Original language	English (US)
Pages (from-to)	91-99
Number of pages	9
Journal	Neurocomputing
Volume	58-60
DOIs	https://doi.org/10.1016/j.neucom.2004.01.028
State	Published - Jun 2004

Bibliographical note

Funding Information:
We thank N.C. Schmitzer-Torbert for generally helpful dialogues. This work was supported by NIH Grant MH68029-01. JCJ was also partly supported by NSF-IGERT training Grant #9870633.

Keywords

Distributed representation
Kernel density estimation (KDE)
Neural ensemble
Reconstruction

Access

10.1016/j.neucom.2004.01.028

OpenUrl availability

Full text

Cite this

@article{7750dfea15ae456e9b37cc62323148c0,

title = "Measuring ensemble consistency without measuring tuning curves",

abstract = "An important question in information processing is the extent to which neural firing patterns remain consistent while processing representations. Transient changes in representational consistency can provide clues to the dynamics of neural processing. We present a generalized framework for measuring the consistency of a neuronal representation that does not require explicit knowledge of the parameters encoded by the ensemble. It requires only neuronal ensembles and a training set of neuronal activity that samples behavioral parameters equally. This will be useful in structures where the behavioral parameters signalled by the neural activity are controversial or unknown.",

keywords = "Distributed representation, Kernel density estimation (KDE), Neural ensemble, Reconstruction",

author = "Jackson, {Jadin C.} and Redish, {A. David}",

note = "Funding Information: We thank N.C. Schmitzer-Torbert for generally helpful dialogues. This work was supported by NIH Grant MH68029-01. JCJ was also partly supported by NSF-IGERT training Grant #9870633.",

year = "2004",

month = jun,

doi = "10.1016/j.neucom.2004.01.028",

language = "English (US)",

volume = "58-60",

pages = "91--99",

journal = "Neurocomputing",

issn = "0925-2312",

publisher = "Elsevier",

}

TY - JOUR

T1 - Measuring ensemble consistency without measuring tuning curves

AU - Jackson, Jadin C.

AU - Redish, A. David

N1 - Funding Information: We thank N.C. Schmitzer-Torbert for generally helpful dialogues. This work was supported by NIH Grant MH68029-01. JCJ was also partly supported by NSF-IGERT training Grant #9870633.

PY - 2004/6

Y1 - 2004/6

N2 - An important question in information processing is the extent to which neural firing patterns remain consistent while processing representations. Transient changes in representational consistency can provide clues to the dynamics of neural processing. We present a generalized framework for measuring the consistency of a neuronal representation that does not require explicit knowledge of the parameters encoded by the ensemble. It requires only neuronal ensembles and a training set of neuronal activity that samples behavioral parameters equally. This will be useful in structures where the behavioral parameters signalled by the neural activity are controversial or unknown.

AB - An important question in information processing is the extent to which neural firing patterns remain consistent while processing representations. Transient changes in representational consistency can provide clues to the dynamics of neural processing. We present a generalized framework for measuring the consistency of a neuronal representation that does not require explicit knowledge of the parameters encoded by the ensemble. It requires only neuronal ensembles and a training set of neuronal activity that samples behavioral parameters equally. This will be useful in structures where the behavioral parameters signalled by the neural activity are controversial or unknown.

KW - Distributed representation

KW - Kernel density estimation (KDE)

KW - Neural ensemble

KW - Reconstruction

UR - http://www.scopus.com/inward/record.url?scp=2542468884&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=2542468884&partnerID=8YFLogxK

U2 - 10.1016/j.neucom.2004.01.028

DO - 10.1016/j.neucom.2004.01.028

M3 - Article

AN - SCOPUS:2542468884

SN - 0925-2312

VL - 58-60

SP - 91

EP - 99

JO - Neurocomputing

JF - Neurocomputing

ER -

Measuring ensemble consistency without measuring tuning curves

Abstract

Bibliographical note

Keywords

Access

OpenUrl availability

Other files and links

Fingerprint

Cite this