Population decoding based on an unfaithful model

S. Wu, H. Nakahara, Noboru Murata, S. Amari

Research output: Chapter in Book/Report/Conference proceedingConference contribution

10 Citations (Scopus)

Abstract

We study a population decoding paradigm in which the maximum likelihood inference is based on an unfaithful decoding model (UMLI). This is usually the case for neural population decoding because the encoding process of the brain is not exactly known, or because a simplified decoding model is preferred for saving computational cost. We consider an unfaithful decoding model which neglects the pair-wise correlation between neuronal activities, and prove that UMLI is asymptotically efficient when the neuronal correlation is uniform or of limited-range. The performance of UMLI is compared with that of the maximum likelihood inference based on a faithful model and that of the center of mass decoding method. It turns out that UMLI has advantages of decreasing the computational complexity remarkablely and maintaining a high-level decoding accuracy at the same time. The effect of correlation on the decoding accuracy is also discussed.

Original languageEnglish
Title of host publicationAdvances in Neural Information Processing Systems
PublisherNeural information processing systems foundation
Pages192-198
Number of pages7
ISBN (Print)0262194503, 9780262194501
Publication statusPublished - 2000
Externally publishedYes
Event13th Annual Neural Information Processing Systems Conference, NIPS 1999 - Denver, CO
Duration: 1999 Nov 291999 Dec 4

Other

Other13th Annual Neural Information Processing Systems Conference, NIPS 1999
CityDenver, CO
Period99/11/2999/12/4

Fingerprint

Decoding
Maximum likelihood
Computational complexity
Brain
Costs

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Information Systems
  • Signal Processing

Cite this

Wu, S., Nakahara, H., Murata, N., & Amari, S. (2000). Population decoding based on an unfaithful model. In Advances in Neural Information Processing Systems (pp. 192-198). Neural information processing systems foundation.

Population decoding based on an unfaithful model. / Wu, S.; Nakahara, H.; Murata, Noboru; Amari, S.

Advances in Neural Information Processing Systems. Neural information processing systems foundation, 2000. p. 192-198.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Wu, S, Nakahara, H, Murata, N & Amari, S 2000, Population decoding based on an unfaithful model. in Advances in Neural Information Processing Systems. Neural information processing systems foundation, pp. 192-198, 13th Annual Neural Information Processing Systems Conference, NIPS 1999, Denver, CO, 99/11/29.
Wu S, Nakahara H, Murata N, Amari S. Population decoding based on an unfaithful model. In Advances in Neural Information Processing Systems. Neural information processing systems foundation. 2000. p. 192-198
Wu, S. ; Nakahara, H. ; Murata, Noboru ; Amari, S. / Population decoding based on an unfaithful model. Advances in Neural Information Processing Systems. Neural information processing systems foundation, 2000. pp. 192-198
@inproceedings{29573c06a8e2400a9c3ed13f74213c7c,
title = "Population decoding based on an unfaithful model",
abstract = "We study a population decoding paradigm in which the maximum likelihood inference is based on an unfaithful decoding model (UMLI). This is usually the case for neural population decoding because the encoding process of the brain is not exactly known, or because a simplified decoding model is preferred for saving computational cost. We consider an unfaithful decoding model which neglects the pair-wise correlation between neuronal activities, and prove that UMLI is asymptotically efficient when the neuronal correlation is uniform or of limited-range. The performance of UMLI is compared with that of the maximum likelihood inference based on a faithful model and that of the center of mass decoding method. It turns out that UMLI has advantages of decreasing the computational complexity remarkablely and maintaining a high-level decoding accuracy at the same time. The effect of correlation on the decoding accuracy is also discussed.",
author = "S. Wu and H. Nakahara and Noboru Murata and S. Amari",
year = "2000",
language = "English",
isbn = "0262194503",
pages = "192--198",
booktitle = "Advances in Neural Information Processing Systems",
publisher = "Neural information processing systems foundation",

}

TY - GEN

T1 - Population decoding based on an unfaithful model

AU - Wu, S.

AU - Nakahara, H.

AU - Murata, Noboru

AU - Amari, S.

PY - 2000

Y1 - 2000

N2 - We study a population decoding paradigm in which the maximum likelihood inference is based on an unfaithful decoding model (UMLI). This is usually the case for neural population decoding because the encoding process of the brain is not exactly known, or because a simplified decoding model is preferred for saving computational cost. We consider an unfaithful decoding model which neglects the pair-wise correlation between neuronal activities, and prove that UMLI is asymptotically efficient when the neuronal correlation is uniform or of limited-range. The performance of UMLI is compared with that of the maximum likelihood inference based on a faithful model and that of the center of mass decoding method. It turns out that UMLI has advantages of decreasing the computational complexity remarkablely and maintaining a high-level decoding accuracy at the same time. The effect of correlation on the decoding accuracy is also discussed.

AB - We study a population decoding paradigm in which the maximum likelihood inference is based on an unfaithful decoding model (UMLI). This is usually the case for neural population decoding because the encoding process of the brain is not exactly known, or because a simplified decoding model is preferred for saving computational cost. We consider an unfaithful decoding model which neglects the pair-wise correlation between neuronal activities, and prove that UMLI is asymptotically efficient when the neuronal correlation is uniform or of limited-range. The performance of UMLI is compared with that of the maximum likelihood inference based on a faithful model and that of the center of mass decoding method. It turns out that UMLI has advantages of decreasing the computational complexity remarkablely and maintaining a high-level decoding accuracy at the same time. The effect of correlation on the decoding accuracy is also discussed.

UR - http://www.scopus.com/inward/record.url?scp=84898973185&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84898973185&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84898973185

SN - 0262194503

SN - 9780262194501

SP - 192

EP - 198

BT - Advances in Neural Information Processing Systems

PB - Neural information processing systems foundation

ER -