An adaptive metric machine for pattern classification

Carlotta Domeniconi, Jing Peng, Dimitrios Gunopulos

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

14 Citations (Scopus)

Abstract

Nearest neighbor classification assumes locally constant class conditional probabilities. This assumption becomes invalid in high dimensions with finite samples due to the curse of dimensionality. Severe bias can be introduced under these conditions when using the nearest neighbor rule. We propose a locally adaptive nearest neighbor classification method to try to minimize bias. We use a Chi-squared distance analysis to compute a flexible metric for producing neighborhoods that are elongated along less relevant feature dimensions and constricted along most influential ones. As a result, the class conditional probabilities tend to be smoother in the modified neighborhoods, whereby better classification performance can be achieved. The efficacy of our method is validated and compared against other techniques using a variety of real world data.

Original languageEnglish
Title of host publicationAdvances in Neural Information Processing Systems 13 - Proceedings of the 2000 Conference, NIPS 2000
PublisherNeural information processing systems foundation
ISBN (Print)0262122413, 9780262122412
StatePublished - 1 Jan 2001
Event14th Annual Neural Information Processing Systems Conference, NIPS 2000 - Denver, CO, United States
Duration: 27 Nov 20002 Dec 2000

Other

Other14th Annual Neural Information Processing Systems Conference, NIPS 2000
CountryUnited States
CityDenver, CO
Period27/11/002/12/00

Fingerprint

Pattern recognition

Cite this

Domeniconi, C., Peng, J., & Gunopulos, D. (2001). An adaptive metric machine for pattern classification. In Advances in Neural Information Processing Systems 13 - Proceedings of the 2000 Conference, NIPS 2000 Neural information processing systems foundation.
Domeniconi, Carlotta ; Peng, Jing ; Gunopulos, Dimitrios. / An adaptive metric machine for pattern classification. Advances in Neural Information Processing Systems 13 - Proceedings of the 2000 Conference, NIPS 2000. Neural information processing systems foundation, 2001.
@inproceedings{c84bc09e332345b496e41810028c1bc6,
title = "An adaptive metric machine for pattern classification",
abstract = "Nearest neighbor classification assumes locally constant class conditional probabilities. This assumption becomes invalid in high dimensions with finite samples due to the curse of dimensionality. Severe bias can be introduced under these conditions when using the nearest neighbor rule. We propose a locally adaptive nearest neighbor classification method to try to minimize bias. We use a Chi-squared distance analysis to compute a flexible metric for producing neighborhoods that are elongated along less relevant feature dimensions and constricted along most influential ones. As a result, the class conditional probabilities tend to be smoother in the modified neighborhoods, whereby better classification performance can be achieved. The efficacy of our method is validated and compared against other techniques using a variety of real world data.",
author = "Carlotta Domeniconi and Jing Peng and Dimitrios Gunopulos",
year = "2001",
month = "1",
day = "1",
language = "English",
isbn = "0262122413",
booktitle = "Advances in Neural Information Processing Systems 13 - Proceedings of the 2000 Conference, NIPS 2000",
publisher = "Neural information processing systems foundation",

}

Domeniconi, C, Peng, J & Gunopulos, D 2001, An adaptive metric machine for pattern classification. in Advances in Neural Information Processing Systems 13 - Proceedings of the 2000 Conference, NIPS 2000. Neural information processing systems foundation, 14th Annual Neural Information Processing Systems Conference, NIPS 2000, Denver, CO, United States, 27/11/00.

An adaptive metric machine for pattern classification. / Domeniconi, Carlotta; Peng, Jing; Gunopulos, Dimitrios.

Advances in Neural Information Processing Systems 13 - Proceedings of the 2000 Conference, NIPS 2000. Neural information processing systems foundation, 2001.

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

TY - GEN

T1 - An adaptive metric machine for pattern classification

AU - Domeniconi, Carlotta

AU - Peng, Jing

AU - Gunopulos, Dimitrios

PY - 2001/1/1

Y1 - 2001/1/1

N2 - Nearest neighbor classification assumes locally constant class conditional probabilities. This assumption becomes invalid in high dimensions with finite samples due to the curse of dimensionality. Severe bias can be introduced under these conditions when using the nearest neighbor rule. We propose a locally adaptive nearest neighbor classification method to try to minimize bias. We use a Chi-squared distance analysis to compute a flexible metric for producing neighborhoods that are elongated along less relevant feature dimensions and constricted along most influential ones. As a result, the class conditional probabilities tend to be smoother in the modified neighborhoods, whereby better classification performance can be achieved. The efficacy of our method is validated and compared against other techniques using a variety of real world data.

AB - Nearest neighbor classification assumes locally constant class conditional probabilities. This assumption becomes invalid in high dimensions with finite samples due to the curse of dimensionality. Severe bias can be introduced under these conditions when using the nearest neighbor rule. We propose a locally adaptive nearest neighbor classification method to try to minimize bias. We use a Chi-squared distance analysis to compute a flexible metric for producing neighborhoods that are elongated along less relevant feature dimensions and constricted along most influential ones. As a result, the class conditional probabilities tend to be smoother in the modified neighborhoods, whereby better classification performance can be achieved. The efficacy of our method is validated and compared against other techniques using a variety of real world data.

UR - http://www.scopus.com/inward/record.url?scp=84898978814&partnerID=8YFLogxK

M3 - Conference contribution

SN - 0262122413

SN - 9780262122412

BT - Advances in Neural Information Processing Systems 13 - Proceedings of the 2000 Conference, NIPS 2000

PB - Neural information processing systems foundation

ER -

Domeniconi C, Peng J, Gunopulos D. An adaptive metric machine for pattern classification. In Advances in Neural Information Processing Systems 13 - Proceedings of the 2000 Conference, NIPS 2000. Neural information processing systems foundation. 2001