Distance covariance analysis

Benjamin R. Cowley, João D. Semedo, Amin Zandvakili, Matthew A. Smith, Adam Kohn, Byron M. Yu

Research output: Contribution to conferencePaper

4 Citations (Scopus)

Abstract

We propose a dimensionality reduction method to identify linear projections that capture interactions between two or more sets of variables. The method, distance covariance analysis (DCA), can detect both linear and nonlinear relationships, and can take dependent variables into account. On previous testbeds and a new testbed that systematically assesses the ability to detect both linear and nonlinear interactions, DCA performs better than or comparable to existing methods, while being one of the fastest methods. To showcase the versatility of DCA, we also applied it to three different neurophysiological datasets.

Original languageEnglish (US)
StatePublished - Jan 1 2017
Event20th International Conference on Artificial Intelligence and Statistics, AISTATS 2017 - Fort Lauderdale, United States
Duration: Apr 20 2017Apr 22 2017

Conference

Conference20th International Conference on Artificial Intelligence and Statistics, AISTATS 2017
CountryUnited States
CityFort Lauderdale
Period4/20/174/22/17

Fingerprint

Testbeds
Testbed
Linear Projection
Nonlinear Interaction
Dimensionality Reduction
Reduction Method
Dependent
Interaction

ASJC Scopus subject areas

  • Artificial Intelligence
  • Statistics and Probability

Cite this

Cowley, B. R., Semedo, J. D., Zandvakili, A., Smith, M. A., Kohn, A., & Yu, B. M. (2017). Distance covariance analysis. Paper presented at 20th International Conference on Artificial Intelligence and Statistics, AISTATS 2017, Fort Lauderdale, United States.

Distance covariance analysis. / Cowley, Benjamin R.; Semedo, João D.; Zandvakili, Amin; Smith, Matthew A.; Kohn, Adam; Yu, Byron M.

2017. Paper presented at 20th International Conference on Artificial Intelligence and Statistics, AISTATS 2017, Fort Lauderdale, United States.

Research output: Contribution to conferencePaper

Cowley, BR, Semedo, JD, Zandvakili, A, Smith, MA, Kohn, A & Yu, BM 2017, 'Distance covariance analysis', Paper presented at 20th International Conference on Artificial Intelligence and Statistics, AISTATS 2017, Fort Lauderdale, United States, 4/20/17 - 4/22/17.
Cowley BR, Semedo JD, Zandvakili A, Smith MA, Kohn A, Yu BM. Distance covariance analysis. 2017. Paper presented at 20th International Conference on Artificial Intelligence and Statistics, AISTATS 2017, Fort Lauderdale, United States.
Cowley, Benjamin R. ; Semedo, João D. ; Zandvakili, Amin ; Smith, Matthew A. ; Kohn, Adam ; Yu, Byron M. / Distance covariance analysis. Paper presented at 20th International Conference on Artificial Intelligence and Statistics, AISTATS 2017, Fort Lauderdale, United States.
@conference{5d6bd9b9d34e442c8512892e355f2d8f,
title = "Distance covariance analysis",
abstract = "We propose a dimensionality reduction method to identify linear projections that capture interactions between two or more sets of variables. The method, distance covariance analysis (DCA), can detect both linear and nonlinear relationships, and can take dependent variables into account. On previous testbeds and a new testbed that systematically assesses the ability to detect both linear and nonlinear interactions, DCA performs better than or comparable to existing methods, while being one of the fastest methods. To showcase the versatility of DCA, we also applied it to three different neurophysiological datasets.",
author = "Cowley, {Benjamin R.} and Semedo, {Jo{\~a}o D.} and Amin Zandvakili and Smith, {Matthew A.} and Adam Kohn and Yu, {Byron M.}",
year = "2017",
month = "1",
day = "1",
language = "English (US)",
note = "20th International Conference on Artificial Intelligence and Statistics, AISTATS 2017 ; Conference date: 20-04-2017 Through 22-04-2017",

}

TY - CONF

T1 - Distance covariance analysis

AU - Cowley, Benjamin R.

AU - Semedo, João D.

AU - Zandvakili, Amin

AU - Smith, Matthew A.

AU - Kohn, Adam

AU - Yu, Byron M.

PY - 2017/1/1

Y1 - 2017/1/1

N2 - We propose a dimensionality reduction method to identify linear projections that capture interactions between two or more sets of variables. The method, distance covariance analysis (DCA), can detect both linear and nonlinear relationships, and can take dependent variables into account. On previous testbeds and a new testbed that systematically assesses the ability to detect both linear and nonlinear interactions, DCA performs better than or comparable to existing methods, while being one of the fastest methods. To showcase the versatility of DCA, we also applied it to three different neurophysiological datasets.

AB - We propose a dimensionality reduction method to identify linear projections that capture interactions between two or more sets of variables. The method, distance covariance analysis (DCA), can detect both linear and nonlinear relationships, and can take dependent variables into account. On previous testbeds and a new testbed that systematically assesses the ability to detect both linear and nonlinear interactions, DCA performs better than or comparable to existing methods, while being one of the fastest methods. To showcase the versatility of DCA, we also applied it to three different neurophysiological datasets.

UR - http://www.scopus.com/inward/record.url?scp=85063382999&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85063382999&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:85063382999

ER -