Abstract

In this paper, we extend previously developed non-parametric bounds on the Bayes risk in binary classification problems to multi-class problems. In comparison with the well-known Bhattacharyya bound which is typically calculated by employing parametric assumptions, the bounds proposed in this paper are directly estimable from data, provably tighter, and more robust to different types of data. We verify the tightness and validity of this bound using an illustrative synthetic example, and further demonstrate its value by incorporating it into a feature selection algorithm which we apply to the real-world problem of distinguishing between different neuro-motor disorders based on sentence-level speech data.

Original languageEnglish (US)
Title of host publication2016 IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP 2016 - Proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages2594-2598
Number of pages5
Volume2016-May
ISBN (Electronic)9781479999880
DOIs
StatePublished - May 18 2016
Event41st IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP 2016 - Shanghai, China
Duration: Mar 20 2016Mar 25 2016

Other

Other41st IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP 2016
CountryChina
CityShanghai
Period3/20/163/25/16

Fingerprint

Feature extraction

Keywords

  • Bayes error rate
  • divergence measures
  • multi-class classification
  • non-parametric estimator

ASJC Scopus subject areas

  • Signal Processing
  • Software
  • Electrical and Electronic Engineering

Cite this

Wisler, A., Berisha, V., Wei, D., Ramamurthy, K., & Spanias, A. (2016). Empirically-estimable multi-class classification bounds. In 2016 IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP 2016 - Proceedings (Vol. 2016-May, pp. 2594-2598). [7472146] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICASSP.2016.7472146

Empirically-estimable multi-class classification bounds. / Wisler, Alan; Berisha, Visar; Wei, Dennis; Ramamurthy, Karthikeyan; Spanias, Andreas.

2016 IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP 2016 - Proceedings. Vol. 2016-May Institute of Electrical and Electronics Engineers Inc., 2016. p. 2594-2598 7472146.

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

Wisler, A, Berisha, V, Wei, D, Ramamurthy, K & Spanias, A 2016, Empirically-estimable multi-class classification bounds. in 2016 IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP 2016 - Proceedings. vol. 2016-May, 7472146, Institute of Electrical and Electronics Engineers Inc., pp. 2594-2598, 41st IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP 2016, Shanghai, China, 3/20/16. https://doi.org/10.1109/ICASSP.2016.7472146
Wisler A, Berisha V, Wei D, Ramamurthy K, Spanias A. Empirically-estimable multi-class classification bounds. In 2016 IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP 2016 - Proceedings. Vol. 2016-May. Institute of Electrical and Electronics Engineers Inc. 2016. p. 2594-2598. 7472146 https://doi.org/10.1109/ICASSP.2016.7472146
Wisler, Alan ; Berisha, Visar ; Wei, Dennis ; Ramamurthy, Karthikeyan ; Spanias, Andreas. / Empirically-estimable multi-class classification bounds. 2016 IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP 2016 - Proceedings. Vol. 2016-May Institute of Electrical and Electronics Engineers Inc., 2016. pp. 2594-2598
@inproceedings{e4cb5c74e3bf4c53b0f13ebaa0ae3c4a,
title = "Empirically-estimable multi-class classification bounds",
abstract = "In this paper, we extend previously developed non-parametric bounds on the Bayes risk in binary classification problems to multi-class problems. In comparison with the well-known Bhattacharyya bound which is typically calculated by employing parametric assumptions, the bounds proposed in this paper are directly estimable from data, provably tighter, and more robust to different types of data. We verify the tightness and validity of this bound using an illustrative synthetic example, and further demonstrate its value by incorporating it into a feature selection algorithm which we apply to the real-world problem of distinguishing between different neuro-motor disorders based on sentence-level speech data.",
keywords = "Bayes error rate, divergence measures, multi-class classification, non-parametric estimator",
author = "Alan Wisler and Visar Berisha and Dennis Wei and Karthikeyan Ramamurthy and Andreas Spanias",
year = "2016",
month = "5",
day = "18",
doi = "10.1109/ICASSP.2016.7472146",
language = "English (US)",
volume = "2016-May",
pages = "2594--2598",
booktitle = "2016 IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP 2016 - Proceedings",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
address = "United States",

}

TY - GEN

T1 - Empirically-estimable multi-class classification bounds

AU - Wisler, Alan

AU - Berisha, Visar

AU - Wei, Dennis

AU - Ramamurthy, Karthikeyan

AU - Spanias, Andreas

PY - 2016/5/18

Y1 - 2016/5/18

N2 - In this paper, we extend previously developed non-parametric bounds on the Bayes risk in binary classification problems to multi-class problems. In comparison with the well-known Bhattacharyya bound which is typically calculated by employing parametric assumptions, the bounds proposed in this paper are directly estimable from data, provably tighter, and more robust to different types of data. We verify the tightness and validity of this bound using an illustrative synthetic example, and further demonstrate its value by incorporating it into a feature selection algorithm which we apply to the real-world problem of distinguishing between different neuro-motor disorders based on sentence-level speech data.

AB - In this paper, we extend previously developed non-parametric bounds on the Bayes risk in binary classification problems to multi-class problems. In comparison with the well-known Bhattacharyya bound which is typically calculated by employing parametric assumptions, the bounds proposed in this paper are directly estimable from data, provably tighter, and more robust to different types of data. We verify the tightness and validity of this bound using an illustrative synthetic example, and further demonstrate its value by incorporating it into a feature selection algorithm which we apply to the real-world problem of distinguishing between different neuro-motor disorders based on sentence-level speech data.

KW - Bayes error rate

KW - divergence measures

KW - multi-class classification

KW - non-parametric estimator

UR - http://www.scopus.com/inward/record.url?scp=84973344431&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84973344431&partnerID=8YFLogxK

U2 - 10.1109/ICASSP.2016.7472146

DO - 10.1109/ICASSP.2016.7472146

M3 - Conference contribution

VL - 2016-May

SP - 2594

EP - 2598

BT - 2016 IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP 2016 - Proceedings

PB - Institute of Electrical and Electronics Engineers Inc.

ER -