Visualizing distortions and recovering topology in continuous projection techniques

Research output: Contribution to journalArticle

44 Citations (Scopus)

Abstract

The visualization of continuous multi-dimensional data based on their projection to a 2-dimensional space is a way to detect visually interesting patterns, as far as the projection provides a faithful image of the original data. In order to evaluate this faithfulness, we propose to visualize any measure associated to a projected datum or to a pair of projected data, by coloring the corresponding Voronoï cell in the projection space. We also define specific measures and show how they allow estimating visually whether some part of the projection is or is not a reliable image of the original manifolds. It also helps to figure out what the original topology of the data is, telling where the high-dimensional manifolds have been torn or glued during the projection. We experiment these techniques with the principal component analysis and the curvilinear component analysis applied to artificial and real databases.

Original languageEnglish
Pages (from-to)1304-1330
Number of pages27
JournalNeurocomputing
Volume70
Issue number7-9
DOIs
Publication statusPublished - Mar 2007
Externally publishedYes

Fingerprint

Coloring
Principal Component Analysis
Principal component analysis
Visualization
Topology
Databases
Experiments

Keywords

  • Continuous projection
  • Delaunay graph
  • Distortion visualization
  • Exploratory data analysis
  • High-dimensional data
  • Topology recovering
  • Uncertainty visualization
  • Voronoï cells

ASJC Scopus subject areas

  • Artificial Intelligence
  • Cellular and Molecular Neuroscience

Cite this

Visualizing distortions and recovering topology in continuous projection techniques. / Aupetit, Michael.

In: Neurocomputing, Vol. 70, No. 7-9, 03.2007, p. 1304-1330.

Research output: Contribution to journalArticle

@article{51b37d2e079346eca6065133ccc94ee0,
title = "Visualizing distortions and recovering topology in continuous projection techniques",
abstract = "The visualization of continuous multi-dimensional data based on their projection to a 2-dimensional space is a way to detect visually interesting patterns, as far as the projection provides a faithful image of the original data. In order to evaluate this faithfulness, we propose to visualize any measure associated to a projected datum or to a pair of projected data, by coloring the corresponding Vorono{\"i} cell in the projection space. We also define specific measures and show how they allow estimating visually whether some part of the projection is or is not a reliable image of the original manifolds. It also helps to figure out what the original topology of the data is, telling where the high-dimensional manifolds have been torn or glued during the projection. We experiment these techniques with the principal component analysis and the curvilinear component analysis applied to artificial and real databases.",
keywords = "Continuous projection, Delaunay graph, Distortion visualization, Exploratory data analysis, High-dimensional data, Topology recovering, Uncertainty visualization, Vorono{\"i} cells",
author = "Michael Aupetit",
year = "2007",
month = "3",
doi = "10.1016/j.neucom.2006.11.018",
language = "English",
volume = "70",
pages = "1304--1330",
journal = "Neurocomputing",
issn = "0925-2312",
publisher = "Elsevier",
number = "7-9",

}

TY - JOUR

T1 - Visualizing distortions and recovering topology in continuous projection techniques

AU - Aupetit, Michael

PY - 2007/3

Y1 - 2007/3

N2 - The visualization of continuous multi-dimensional data based on their projection to a 2-dimensional space is a way to detect visually interesting patterns, as far as the projection provides a faithful image of the original data. In order to evaluate this faithfulness, we propose to visualize any measure associated to a projected datum or to a pair of projected data, by coloring the corresponding Voronoï cell in the projection space. We also define specific measures and show how they allow estimating visually whether some part of the projection is or is not a reliable image of the original manifolds. It also helps to figure out what the original topology of the data is, telling where the high-dimensional manifolds have been torn or glued during the projection. We experiment these techniques with the principal component analysis and the curvilinear component analysis applied to artificial and real databases.

AB - The visualization of continuous multi-dimensional data based on their projection to a 2-dimensional space is a way to detect visually interesting patterns, as far as the projection provides a faithful image of the original data. In order to evaluate this faithfulness, we propose to visualize any measure associated to a projected datum or to a pair of projected data, by coloring the corresponding Voronoï cell in the projection space. We also define specific measures and show how they allow estimating visually whether some part of the projection is or is not a reliable image of the original manifolds. It also helps to figure out what the original topology of the data is, telling where the high-dimensional manifolds have been torn or glued during the projection. We experiment these techniques with the principal component analysis and the curvilinear component analysis applied to artificial and real databases.

KW - Continuous projection

KW - Delaunay graph

KW - Distortion visualization

KW - Exploratory data analysis

KW - High-dimensional data

KW - Topology recovering

KW - Uncertainty visualization

KW - Voronoï cells

UR - http://www.scopus.com/inward/record.url?scp=33847360579&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33847360579&partnerID=8YFLogxK

U2 - 10.1016/j.neucom.2006.11.018

DO - 10.1016/j.neucom.2006.11.018

M3 - Article

VL - 70

SP - 1304

EP - 1330

JO - Neurocomputing

JF - Neurocomputing

SN - 0925-2312

IS - 7-9

ER -