Flattening the density gradient for eliminating spatial centrality to reduce hubness

Kazuo Hara, Ikumi Suzuki, Kei Kobayashi, Kenji Fukumizu, Milǒs Radovanovíc

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

3 Citations (Scopus)

Abstract

Spatial centrality, whereby samples closer to the center of a dataset tend to be closer to all other samples, is regarded as one source of hubness. Hubness is well known to degrade k-nearest-neighbor (k-NN) classification. Spatial centrality can be removed by centering, i.e., shifting the origin to the global center of the dataset, in cases where inner product similarity is used. However, when Euclidean distance is used, centering has no effect on spatial centrality because the distance between the samples is the same before and after centering. As described in this paper, we propose a solution for the hubness problem when Euclidean distance is considered. We provide a theoretical explanation to demonstrate how the solution eliminates spatial centrality and reduces hubness. We then present some discussion of the reason the proposed solution works, from a viewpoint of density gradient, which is regarded as the origin of spatial centrality and hubness. We demonstrate that the solution corresponds to flattening the density gradient. Using real-world datasets, we demonstrate that the proposed method improves k-NN classification performance and outperforms an existing hub-reduction method.

Original languageEnglish
Title of host publication30th AAAI Conference on Artificial Intelligence, AAAI 2016
PublisherAAAI press
Pages1659-1665
Number of pages7
ISBN (Electronic)9781577357605
Publication statusPublished - 2016
Externally publishedYes
Event30th AAAI Conference on Artificial Intelligence, AAAI 2016 - Phoenix, United States
Duration: 2016 Feb 122016 Feb 17

Other

Other30th AAAI Conference on Artificial Intelligence, AAAI 2016
CountryUnited States
CityPhoenix
Period16/2/1216/2/17

ASJC Scopus subject areas

  • Artificial Intelligence

Cite this

Hara, K., Suzuki, I., Kobayashi, K., Fukumizu, K., & Radovanovíc, M. (2016). Flattening the density gradient for eliminating spatial centrality to reduce hubness. In 30th AAAI Conference on Artificial Intelligence, AAAI 2016 (pp. 1659-1665). AAAI press.

Flattening the density gradient for eliminating spatial centrality to reduce hubness. / Hara, Kazuo; Suzuki, Ikumi; Kobayashi, Kei; Fukumizu, Kenji; Radovanovíc, Milǒs.

30th AAAI Conference on Artificial Intelligence, AAAI 2016. AAAI press, 2016. p. 1659-1665.

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

Hara, K, Suzuki, I, Kobayashi, K, Fukumizu, K & Radovanovíc, M 2016, Flattening the density gradient for eliminating spatial centrality to reduce hubness. in 30th AAAI Conference on Artificial Intelligence, AAAI 2016. AAAI press, pp. 1659-1665, 30th AAAI Conference on Artificial Intelligence, AAAI 2016, Phoenix, United States, 16/2/12.
Hara K, Suzuki I, Kobayashi K, Fukumizu K, Radovanovíc M. Flattening the density gradient for eliminating spatial centrality to reduce hubness. In 30th AAAI Conference on Artificial Intelligence, AAAI 2016. AAAI press. 2016. p. 1659-1665
Hara, Kazuo ; Suzuki, Ikumi ; Kobayashi, Kei ; Fukumizu, Kenji ; Radovanovíc, Milǒs. / Flattening the density gradient for eliminating spatial centrality to reduce hubness. 30th AAAI Conference on Artificial Intelligence, AAAI 2016. AAAI press, 2016. pp. 1659-1665
@inproceedings{94959ccc36db4f139d73b1a6d880ad30,
title = "Flattening the density gradient for eliminating spatial centrality to reduce hubness",
abstract = "Spatial centrality, whereby samples closer to the center of a dataset tend to be closer to all other samples, is regarded as one source of hubness. Hubness is well known to degrade k-nearest-neighbor (k-NN) classification. Spatial centrality can be removed by centering, i.e., shifting the origin to the global center of the dataset, in cases where inner product similarity is used. However, when Euclidean distance is used, centering has no effect on spatial centrality because the distance between the samples is the same before and after centering. As described in this paper, we propose a solution for the hubness problem when Euclidean distance is considered. We provide a theoretical explanation to demonstrate how the solution eliminates spatial centrality and reduces hubness. We then present some discussion of the reason the proposed solution works, from a viewpoint of density gradient, which is regarded as the origin of spatial centrality and hubness. We demonstrate that the solution corresponds to flattening the density gradient. Using real-world datasets, we demonstrate that the proposed method improves k-NN classification performance and outperforms an existing hub-reduction method.",
author = "Kazuo Hara and Ikumi Suzuki and Kei Kobayashi and Kenji Fukumizu and Milǒs Radovanov{\'i}c",
year = "2016",
language = "English",
pages = "1659--1665",
booktitle = "30th AAAI Conference on Artificial Intelligence, AAAI 2016",
publisher = "AAAI press",

}

TY - GEN

T1 - Flattening the density gradient for eliminating spatial centrality to reduce hubness

AU - Hara, Kazuo

AU - Suzuki, Ikumi

AU - Kobayashi, Kei

AU - Fukumizu, Kenji

AU - Radovanovíc, Milǒs

PY - 2016

Y1 - 2016

N2 - Spatial centrality, whereby samples closer to the center of a dataset tend to be closer to all other samples, is regarded as one source of hubness. Hubness is well known to degrade k-nearest-neighbor (k-NN) classification. Spatial centrality can be removed by centering, i.e., shifting the origin to the global center of the dataset, in cases where inner product similarity is used. However, when Euclidean distance is used, centering has no effect on spatial centrality because the distance between the samples is the same before and after centering. As described in this paper, we propose a solution for the hubness problem when Euclidean distance is considered. We provide a theoretical explanation to demonstrate how the solution eliminates spatial centrality and reduces hubness. We then present some discussion of the reason the proposed solution works, from a viewpoint of density gradient, which is regarded as the origin of spatial centrality and hubness. We demonstrate that the solution corresponds to flattening the density gradient. Using real-world datasets, we demonstrate that the proposed method improves k-NN classification performance and outperforms an existing hub-reduction method.

AB - Spatial centrality, whereby samples closer to the center of a dataset tend to be closer to all other samples, is regarded as one source of hubness. Hubness is well known to degrade k-nearest-neighbor (k-NN) classification. Spatial centrality can be removed by centering, i.e., shifting the origin to the global center of the dataset, in cases where inner product similarity is used. However, when Euclidean distance is used, centering has no effect on spatial centrality because the distance between the samples is the same before and after centering. As described in this paper, we propose a solution for the hubness problem when Euclidean distance is considered. We provide a theoretical explanation to demonstrate how the solution eliminates spatial centrality and reduces hubness. We then present some discussion of the reason the proposed solution works, from a viewpoint of density gradient, which is regarded as the origin of spatial centrality and hubness. We demonstrate that the solution corresponds to flattening the density gradient. Using real-world datasets, we demonstrate that the proposed method improves k-NN classification performance and outperforms an existing hub-reduction method.

UR - http://www.scopus.com/inward/record.url?scp=85007152062&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85007152062&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:85007152062

SP - 1659

EP - 1665

BT - 30th AAAI Conference on Artificial Intelligence, AAAI 2016

PB - AAAI press

ER -