Superiority of nonlinear mapping in decoding multiple single-unit neuronal spike trains: A simulation study

Kyung Hwan Kim, Sung Shin Kim, Sung June Kim

Research output: Contribution to journalArticle

22 Citations (Scopus)

Abstract

One of the most important building blocks of the brain-machine interface (BMI) based on neuronal spike trains is the decoding algorithm, a computational method for the reconstruction of desired information from spike trains. Previous studies have reported that a simple linear filter is effective for this purpose and that no noteworthy gain is achieved from the use of nonlinear algorithms. In order to test this premise, we designed several decoding algorithms based on the linear filter, and two nonlinear mapping algorithms using multilayer perceptron (MLP) and support vector machine regression (SVR). Their performances were assessed using multiple neuronal spike trains generated by a biophysical neuron model and by a directional tuning model of the primary motor cortex. The performances of the nonlinear algorithms, in general, were superior. The advantages of using nonlinear algorithms were more profound for cases where false-positive/negative errors occurred in spike trains. When the MLPs were trained using trial-and-error, they often showed disappointing performance comparable to that of the linear filter. The nonlinear SVR showed the highest performance, and this may be due to the superiority of SVR in training and generalization.

Original languageEnglish
Pages (from-to)202-211
Number of pages10
JournalJournal of Neuroscience Methods
Volume150
Issue number2
DOIs
Publication statusPublished - 2006 Jan 30

Fingerprint

Brain-Computer Interfaces
Neural Networks (Computer)
Motor Cortex
Neurons
Support Vector Machine

All Science Journal Classification (ASJC) codes

  • Neuroscience(all)

Cite this

@article{78e205c2fd9847ba82614f1f3accc04c,
title = "Superiority of nonlinear mapping in decoding multiple single-unit neuronal spike trains: A simulation study",
abstract = "One of the most important building blocks of the brain-machine interface (BMI) based on neuronal spike trains is the decoding algorithm, a computational method for the reconstruction of desired information from spike trains. Previous studies have reported that a simple linear filter is effective for this purpose and that no noteworthy gain is achieved from the use of nonlinear algorithms. In order to test this premise, we designed several decoding algorithms based on the linear filter, and two nonlinear mapping algorithms using multilayer perceptron (MLP) and support vector machine regression (SVR). Their performances were assessed using multiple neuronal spike trains generated by a biophysical neuron model and by a directional tuning model of the primary motor cortex. The performances of the nonlinear algorithms, in general, were superior. The advantages of using nonlinear algorithms were more profound for cases where false-positive/negative errors occurred in spike trains. When the MLPs were trained using trial-and-error, they often showed disappointing performance comparable to that of the linear filter. The nonlinear SVR showed the highest performance, and this may be due to the superiority of SVR in training and generalization.",
author = "Kim, {Kyung Hwan} and Kim, {Sung Shin} and Kim, {Sung June}",
year = "2006",
month = "1",
day = "30",
doi = "10.1016/j.jneumeth.2005.06.015",
language = "English",
volume = "150",
pages = "202--211",
journal = "Journal of Neuroscience Methods",
issn = "0165-0270",
publisher = "Elsevier",
number = "2",

}

Superiority of nonlinear mapping in decoding multiple single-unit neuronal spike trains : A simulation study. / Kim, Kyung Hwan; Kim, Sung Shin; Kim, Sung June.

In: Journal of Neuroscience Methods, Vol. 150, No. 2, 30.01.2006, p. 202-211.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Superiority of nonlinear mapping in decoding multiple single-unit neuronal spike trains

T2 - A simulation study

AU - Kim, Kyung Hwan

AU - Kim, Sung Shin

AU - Kim, Sung June

PY - 2006/1/30

Y1 - 2006/1/30

N2 - One of the most important building blocks of the brain-machine interface (BMI) based on neuronal spike trains is the decoding algorithm, a computational method for the reconstruction of desired information from spike trains. Previous studies have reported that a simple linear filter is effective for this purpose and that no noteworthy gain is achieved from the use of nonlinear algorithms. In order to test this premise, we designed several decoding algorithms based on the linear filter, and two nonlinear mapping algorithms using multilayer perceptron (MLP) and support vector machine regression (SVR). Their performances were assessed using multiple neuronal spike trains generated by a biophysical neuron model and by a directional tuning model of the primary motor cortex. The performances of the nonlinear algorithms, in general, were superior. The advantages of using nonlinear algorithms were more profound for cases where false-positive/negative errors occurred in spike trains. When the MLPs were trained using trial-and-error, they often showed disappointing performance comparable to that of the linear filter. The nonlinear SVR showed the highest performance, and this may be due to the superiority of SVR in training and generalization.

AB - One of the most important building blocks of the brain-machine interface (BMI) based on neuronal spike trains is the decoding algorithm, a computational method for the reconstruction of desired information from spike trains. Previous studies have reported that a simple linear filter is effective for this purpose and that no noteworthy gain is achieved from the use of nonlinear algorithms. In order to test this premise, we designed several decoding algorithms based on the linear filter, and two nonlinear mapping algorithms using multilayer perceptron (MLP) and support vector machine regression (SVR). Their performances were assessed using multiple neuronal spike trains generated by a biophysical neuron model and by a directional tuning model of the primary motor cortex. The performances of the nonlinear algorithms, in general, were superior. The advantages of using nonlinear algorithms were more profound for cases where false-positive/negative errors occurred in spike trains. When the MLPs were trained using trial-and-error, they often showed disappointing performance comparable to that of the linear filter. The nonlinear SVR showed the highest performance, and this may be due to the superiority of SVR in training and generalization.

UR - http://www.scopus.com/inward/record.url?scp=31344442525&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=31344442525&partnerID=8YFLogxK

U2 - 10.1016/j.jneumeth.2005.06.015

DO - 10.1016/j.jneumeth.2005.06.015

M3 - Article

C2 - 16099513

AN - SCOPUS:31344442525

VL - 150

SP - 202

EP - 211

JO - Journal of Neuroscience Methods

JF - Journal of Neuroscience Methods

SN - 0165-0270

IS - 2

ER -