Explaining Neighborhood Preservation for Multidimensional Projections

Martins, Rafael Messias; Minghim, Rosane; Telea, Alexandru C.

Explaining Neighborhood Preservation for Multidimensional Projections

Files

007-014.pdf (1.15 MB)

Date

2015

Authors

Martins, Rafael Messias
Minghim, Rosane
Telea, Alexandru C.

Publisher

The Eurographics Association

Abstract

Dimensionality reduction techniques are the tools of choice for exploring high-dimensional datasets by means of low-dimensional projections. However, even state-of-the-art projection methods fail, up to various degrees, in perfectly preserving the structure of the data, expressed in terms of inter-point distances and point neighborhoods. To support better interpretation of a projection, we propose several metrics for quantifying errors related to neighborhood preservation. Next, we propose a number of visualizations that allow users to explore and explain the quality of neighborhood preservation at different scales, captured by the aforementioned error metrics.We demonstrate our exploratory views on three real-world datasets and two state-of-the-art multidimensional projection techniques.

        @inproceedings{10.2312:cgvc.20151234
,
booktitle = {Computer Graphics and Visual Computing (CGVC)
},
editor = {Rita Borgo and Cagatay Turkay
},
title = {{Explaining Neighborhood Preservation for Multidimensional Projections
}},
author = {Martins, Rafael Messias and 
Minghim, Rosane and 
Telea, Alexandru C.
},
year = {2015
},
publisher = {The Eurographics Association
},
ISBN = {978-3-905674-94-1
},
DOI = {10.2312/cgvc.20151234
}
}