Learning multiscale neural metrics via entropy minimization

Austin J. Brockmeier, Luis G.Sanchez Giraldo, John S. Choi, Joseph T. Francis, Jose C. Principe

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Scopus citations

Abstract

In order to judiciously compare neural responses between repeated trials or stimuli, a well-suited distance metric is necessary. With multi-electrode recordings, a neural response is a spatiotemporal pattern, but not all of the dimensions of space and time should be treated equally. In order to understand which dimensions of the input are more discriminative and to improve the classification performance, we propose a metric-learning approach that can be used across scales. This extends previous work that used a linear projection into lower dimensional space; here, multiscale metrics or kernels are learned as the weighted combinations of different metrics or kernels on each of the neural response's dimensions. Preliminary results are explored on a cortical recording of a rat during a tactile stimulation experiment. Metrics on both local field potential and spiking data are explored. The learned weights reveal important dimensions of the response, and the learned metrics improve nearest-neighbor classification performance.

Original languageEnglish
Title of host publication2013 6th International IEEE EMBS Conference on Neural Engineering, NER 2013
Pages247-250
Number of pages4
DOIs
StatePublished - 2013
Event2013 6th International IEEE EMBS Conference on Neural Engineering, NER 2013 - San Diego, CA, United States
Duration: Nov 6 2013Nov 8 2013

Publication series

NameInternational IEEE/EMBS Conference on Neural Engineering, NER
ISSN (Print)1948-3546
ISSN (Electronic)1948-3554

Conference

Conference2013 6th International IEEE EMBS Conference on Neural Engineering, NER 2013
Country/TerritoryUnited States
CitySan Diego, CA
Period11/6/1311/8/13

ASJC Scopus subject areas

  • Artificial Intelligence
  • Mechanical Engineering

Fingerprint

Dive into the research topics of 'Learning multiscale neural metrics via entropy minimization'. Together they form a unique fingerprint.

Cite this