Home > Research > Publications & Outputs > A new type of distance metric and its use for c...

Electronic data

  • DirectionAwareDistanceeClustering_evos

    Rights statement: The final publication is available at Springer via http://dx.doi.org/10.1007/s12530-017-9195-7

    Accepted author manuscript, 626 KB, PDF document

    Available under license: CC BY-NC: Creative Commons Attribution-NonCommercial 4.0 International License

Links

Text available via DOI:

View graph of relations

A new type of distance metric and its use for clustering

Research output: Contribution to Journal/MagazineJournal articlepeer-review

Published

Standard

A new type of distance metric and its use for clustering. / Gu, Xiaowei; Angelov, Plamen Parvanov; Kangin, Dmitry et al.
In: Evolving Systems, Vol. 8, No. 3, 09.2017, p. 167-177.

Research output: Contribution to Journal/MagazineJournal articlepeer-review

Harvard

Gu, X, Angelov, PP, Kangin, D & Principe, J 2017, 'A new type of distance metric and its use for clustering', Evolving Systems, vol. 8, no. 3, pp. 167-177. https://doi.org/10.1007/s12530-017-9195-7

APA

Vancouver

Gu X, Angelov PP, Kangin D, Principe J. A new type of distance metric and its use for clustering. Evolving Systems. 2017 Sept;8(3):167-177. Epub 2017 Jul 26. doi: 10.1007/s12530-017-9195-7

Author

Gu, Xiaowei ; Angelov, Plamen Parvanov ; Kangin, Dmitry et al. / A new type of distance metric and its use for clustering. In: Evolving Systems. 2017 ; Vol. 8, No. 3. pp. 167-177.

Bibtex

@article{0a6c06e40d0a46ba9900d4fe79d970a8,
title = "A new type of distance metric and its use for clustering",
abstract = "In order to address high dimensional problems, a new {\textquoteleft}direction-aware{\textquoteright} metric is introduced in this paper. This new distance is a combination of two components: i) the traditional Euclidean distance and ii) an angular/directional divergence, derived from the cosine similarity. The newly introduced metric combines the advantages of the Euclidean metric and cosine similarity, and is defined over the Euclidean space domain. Thus, it is able to take the advantage from both spaces, while preserving the Euclidean space domain. The direction-aware distance has wide range of applicability and can be used as an alternative distance measure for various traditional clustering approaches to enhance their ability of handling high dimensional problems. A new evolving clustering algorithm using the proposed distance is also proposed in this paper. Numerical examples with benchmark datasets reveal that the direction-aware distance can effectively improve the clustering quality of the k-means algorithm for high dimensional problems and demonstrate the proposed evolving clustering algorithm to be an effective tool for high dimensional data streams processing.",
keywords = "cosine similarity, distance metric, metric space, clustering, high dimensional data streams processing.",
author = "Xiaowei Gu and Angelov, {Plamen Parvanov} and Dmitry Kangin and Jose Principe",
note = "The final publication is available at Springer via http://dx.doi.org/10.1007/s12530-017-9195-7",
year = "2017",
month = sep,
doi = "10.1007/s12530-017-9195-7",
language = "English",
volume = "8",
pages = "167--177",
journal = "Evolving Systems",
issn = "1868-6478",
publisher = "Springer Verlag",
number = "3",

}

RIS

TY - JOUR

T1 - A new type of distance metric and its use for clustering

AU - Gu, Xiaowei

AU - Angelov, Plamen Parvanov

AU - Kangin, Dmitry

AU - Principe, Jose

N1 - The final publication is available at Springer via http://dx.doi.org/10.1007/s12530-017-9195-7

PY - 2017/9

Y1 - 2017/9

N2 - In order to address high dimensional problems, a new ‘direction-aware’ metric is introduced in this paper. This new distance is a combination of two components: i) the traditional Euclidean distance and ii) an angular/directional divergence, derived from the cosine similarity. The newly introduced metric combines the advantages of the Euclidean metric and cosine similarity, and is defined over the Euclidean space domain. Thus, it is able to take the advantage from both spaces, while preserving the Euclidean space domain. The direction-aware distance has wide range of applicability and can be used as an alternative distance measure for various traditional clustering approaches to enhance their ability of handling high dimensional problems. A new evolving clustering algorithm using the proposed distance is also proposed in this paper. Numerical examples with benchmark datasets reveal that the direction-aware distance can effectively improve the clustering quality of the k-means algorithm for high dimensional problems and demonstrate the proposed evolving clustering algorithm to be an effective tool for high dimensional data streams processing.

AB - In order to address high dimensional problems, a new ‘direction-aware’ metric is introduced in this paper. This new distance is a combination of two components: i) the traditional Euclidean distance and ii) an angular/directional divergence, derived from the cosine similarity. The newly introduced metric combines the advantages of the Euclidean metric and cosine similarity, and is defined over the Euclidean space domain. Thus, it is able to take the advantage from both spaces, while preserving the Euclidean space domain. The direction-aware distance has wide range of applicability and can be used as an alternative distance measure for various traditional clustering approaches to enhance their ability of handling high dimensional problems. A new evolving clustering algorithm using the proposed distance is also proposed in this paper. Numerical examples with benchmark datasets reveal that the direction-aware distance can effectively improve the clustering quality of the k-means algorithm for high dimensional problems and demonstrate the proposed evolving clustering algorithm to be an effective tool for high dimensional data streams processing.

KW - cosine similarity

KW - distance metric

KW - metric space

KW - clustering

KW - high dimensional data streams processing.

U2 - 10.1007/s12530-017-9195-7

DO - 10.1007/s12530-017-9195-7

M3 - Journal article

VL - 8

SP - 167

EP - 177

JO - Evolving Systems

JF - Evolving Systems

SN - 1868-6478

IS - 3

ER -