LoCH: A neighborhood-based multidimensional projection technique for high-dimensional sparse spaces

Publikation: Beiträge in ZeitschriftenZeitschriftenaufsätzeForschungbegutachtet

Standard

LoCH: A neighborhood-based multidimensional projection technique for high-dimensional sparse spaces. / Fadel, Samuel G.; Fatore, Francisco M.; Duarte, Felipe S.L.G. et al.
in: Neurocomputing, Jahrgang 150, Nr. Part B, 20.02.2015, S. 546-556.

Publikation: Beiträge in ZeitschriftenZeitschriftenaufsätzeForschungbegutachtet

Harvard

APA

Vancouver

Fadel SG, Fatore FM, Duarte FSLG, Paulovich FV. LoCH: A neighborhood-based multidimensional projection technique for high-dimensional sparse spaces. Neurocomputing. 2015 Feb 20;150(Part B):546-556. doi: 10.1016/j.neucom.2014.07.071

Bibtex

@article{339ef76be8c54785bf1be9859df47378,
title = "LoCH: A neighborhood-based multidimensional projection technique for high-dimensional sparse spaces",
abstract = "On the last few years multidimensional projection techniques have advanced towards defining faster and user-centered approaches. However, most of existing methods are designed as generic tools without considering particular features of the data under processing, such as the distance distribution when the data is embedded into a certain metric space. In this paper we split the projection techniques into two groups, global and local techniques, conduct an analysis of them, and present a novel local technique specially designed for projecting heavy tail distance distributions, such as the one produced by high-dimensional sparse spaces. This novel approach, called Local Convex Hull (LoCH), relies on an iterative process that seeks to place each point close to the convex hull of its nearest neighbors. The accuracy, in terms of neighborhood preservation, is confirmed by a set of comparisons and tests, showing that LoCH is capable of successfully segregating groups of similar instances embedded in high-dimensional sparse spaces and of defining the borders between them, significantly better than most projection techniques.",
keywords = "High-dimensional sparse space, Local multidimensional projection, Visual data mining, Business informatics",
author = "Fadel, {Samuel G.} and Fatore, {Francisco M.} and Duarte, {Felipe S.L.G.} and Paulovich, {Fernando V.}",
year = "2015",
month = feb,
day = "20",
doi = "10.1016/j.neucom.2014.07.071",
language = "English",
volume = "150",
pages = "546--556",
journal = "Neurocomputing",
issn = "0925-2312",
publisher = "Elsevier B.V.",
number = "Part B",

}

RIS

TY - JOUR

T1 - LoCH

T2 - A neighborhood-based multidimensional projection technique for high-dimensional sparse spaces

AU - Fadel, Samuel G.

AU - Fatore, Francisco M.

AU - Duarte, Felipe S.L.G.

AU - Paulovich, Fernando V.

PY - 2015/2/20

Y1 - 2015/2/20

N2 - On the last few years multidimensional projection techniques have advanced towards defining faster and user-centered approaches. However, most of existing methods are designed as generic tools without considering particular features of the data under processing, such as the distance distribution when the data is embedded into a certain metric space. In this paper we split the projection techniques into two groups, global and local techniques, conduct an analysis of them, and present a novel local technique specially designed for projecting heavy tail distance distributions, such as the one produced by high-dimensional sparse spaces. This novel approach, called Local Convex Hull (LoCH), relies on an iterative process that seeks to place each point close to the convex hull of its nearest neighbors. The accuracy, in terms of neighborhood preservation, is confirmed by a set of comparisons and tests, showing that LoCH is capable of successfully segregating groups of similar instances embedded in high-dimensional sparse spaces and of defining the borders between them, significantly better than most projection techniques.

AB - On the last few years multidimensional projection techniques have advanced towards defining faster and user-centered approaches. However, most of existing methods are designed as generic tools without considering particular features of the data under processing, such as the distance distribution when the data is embedded into a certain metric space. In this paper we split the projection techniques into two groups, global and local techniques, conduct an analysis of them, and present a novel local technique specially designed for projecting heavy tail distance distributions, such as the one produced by high-dimensional sparse spaces. This novel approach, called Local Convex Hull (LoCH), relies on an iterative process that seeks to place each point close to the convex hull of its nearest neighbors. The accuracy, in terms of neighborhood preservation, is confirmed by a set of comparisons and tests, showing that LoCH is capable of successfully segregating groups of similar instances embedded in high-dimensional sparse spaces and of defining the borders between them, significantly better than most projection techniques.

KW - High-dimensional sparse space

KW - Local multidimensional projection

KW - Visual data mining

KW - Business informatics

UR - http://www.scopus.com/inward/record.url?scp=84922676246&partnerID=8YFLogxK

U2 - 10.1016/j.neucom.2014.07.071

DO - 10.1016/j.neucom.2014.07.071

M3 - Journal articles

AN - SCOPUS:84922676246

VL - 150

SP - 546

EP - 556

JO - Neurocomputing

JF - Neurocomputing

SN - 0925-2312

IS - Part B

ER -

DOI