Visual pattern discovery using random projections

Anushka Anand; Leland Wilkinson; Tuan Nhon Dang

doi:10.1109/VAST.2012.6400490

Visual pattern discovery using random projections

Anushka Anand, Leland Wilkinson, Tuan Nhon Dang

Computer Science

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

30 Scopus citations

Abstract

An essential element of exploratory data analysis is the use of revealing low-dimensional projections of high-dimensional data. Projection Pursuit has been an effective method for finding interesting low-dimensional projections of multidimensional spaces by optimizing a score function called a projection pursuit index. However, the technique is not scalable to high-dimensional spaces. Here, we introduce a novel method for discovering noteworthy views of high-dimensional data spaces by using binning and random projections. We define score functions, akin to projection pursuit indices, that characterize visual patterns of the low-dimensional projections that constitute feature subspaces. We also describe an analytic, multivariate visualization platform based on this algorithm that is scalable to extremely large problems.

Original language	English
Title of host publication	IEEE Conference on Visual Analytics Science and Technology 2012, VAST 2012 - Proceedings
Pages	43-52
Number of pages	10
DOIs	https://doi.org/10.1109/VAST.2012.6400490
State	Published - 2012
Event	2012 IEEE Conference on Visual Analytics Science and Technology, VAST 2012 - Seattle, WA, United States Duration: Oct 14 2012 → Oct 19 2012

Publication series

Name	IEEE Conference on Visual Analytics Science and Technology 2012, VAST 2012 - Proceedings

Conference

Conference	2012 IEEE Conference on Visual Analytics Science and Technology, VAST 2012
Country/Territory	United States
City	Seattle, WA
Period	10/14/12 → 10/19/12

Keywords

High-dimensional Data
Random Projections

Access to Document

10.1109/VAST.2012.6400490

Cite this

@inproceedings{6f01ec2a90534f29b28f62fbd3b8f4c7,

title = "Visual pattern discovery using random projections",

abstract = "An essential element of exploratory data analysis is the use of revealing low-dimensional projections of high-dimensional data. Projection Pursuit has been an effective method for finding interesting low-dimensional projections of multidimensional spaces by optimizing a score function called a projection pursuit index. However, the technique is not scalable to high-dimensional spaces. Here, we introduce a novel method for discovering noteworthy views of high-dimensional data spaces by using binning and random projections. We define score functions, akin to projection pursuit indices, that characterize visual patterns of the low-dimensional projections that constitute feature subspaces. We also describe an analytic, multivariate visualization platform based on this algorithm that is scalable to extremely large problems.",

keywords = "High-dimensional Data, Random Projections",

author = "Anushka Anand and Leland Wilkinson and Dang, {Tuan Nhon}",

year = "2012",

doi = "10.1109/VAST.2012.6400490",

language = "English",

isbn = "9781467347532",

series = "IEEE Conference on Visual Analytics Science and Technology 2012, VAST 2012 - Proceedings",

pages = "43--52",

booktitle = "IEEE Conference on Visual Analytics Science and Technology 2012, VAST 2012 - Proceedings",

note = "null ; Conference date: 14-10-2012 Through 19-10-2012",

}

Anand, A, Wilkinson, L & Dang, TN 2012, Visual pattern discovery using random projections. in IEEE Conference on Visual Analytics Science and Technology 2012, VAST 2012 - Proceedings., 6400490, IEEE Conference on Visual Analytics Science and Technology 2012, VAST 2012 - Proceedings, pp. 43-52, 2012 IEEE Conference on Visual Analytics Science and Technology, VAST 2012, Seattle, WA, United States, 10/14/12. https://doi.org/10.1109/VAST.2012.6400490

Visual pattern discovery using random projections. / Anand, Anushka; Wilkinson, Leland; Dang, Tuan Nhon.

IEEE Conference on Visual Analytics Science and Technology 2012, VAST 2012 - Proceedings. 2012. p. 43-52 6400490 (IEEE Conference on Visual Analytics Science and Technology 2012, VAST 2012 - Proceedings).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Visual pattern discovery using random projections

AU - Anand, Anushka

AU - Wilkinson, Leland

AU - Dang, Tuan Nhon

PY - 2012

Y1 - 2012

N2 - An essential element of exploratory data analysis is the use of revealing low-dimensional projections of high-dimensional data. Projection Pursuit has been an effective method for finding interesting low-dimensional projections of multidimensional spaces by optimizing a score function called a projection pursuit index. However, the technique is not scalable to high-dimensional spaces. Here, we introduce a novel method for discovering noteworthy views of high-dimensional data spaces by using binning and random projections. We define score functions, akin to projection pursuit indices, that characterize visual patterns of the low-dimensional projections that constitute feature subspaces. We also describe an analytic, multivariate visualization platform based on this algorithm that is scalable to extremely large problems.

AB - An essential element of exploratory data analysis is the use of revealing low-dimensional projections of high-dimensional data. Projection Pursuit has been an effective method for finding interesting low-dimensional projections of multidimensional spaces by optimizing a score function called a projection pursuit index. However, the technique is not scalable to high-dimensional spaces. Here, we introduce a novel method for discovering noteworthy views of high-dimensional data spaces by using binning and random projections. We define score functions, akin to projection pursuit indices, that characterize visual patterns of the low-dimensional projections that constitute feature subspaces. We also describe an analytic, multivariate visualization platform based on this algorithm that is scalable to extremely large problems.

KW - High-dimensional Data

KW - Random Projections

UR - http://www.scopus.com/inward/record.url?scp=84872963385&partnerID=8YFLogxK

U2 - 10.1109/VAST.2012.6400490

DO - 10.1109/VAST.2012.6400490

M3 - Conference contribution

AN - SCOPUS:84872963385

SN - 9781467347532

T3 - IEEE Conference on Visual Analytics Science and Technology 2012, VAST 2012 - Proceedings

SP - 43

EP - 52

BT - IEEE Conference on Visual Analytics Science and Technology 2012, VAST 2012 - Proceedings

Y2 - 14 October 2012 through 19 October 2012

ER -

Visual pattern discovery using random projections

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this