Kleczynski, M.
and Kearsley, A.
(2025),
Topological Initialization for Multidimensional Scaling, Technical Note (NIST TN), National Institute of Standards and Technology, Gaithersburg, MD, [online], https://doi.org/10.6028/NIST.TN.2349, https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=960027
(Accessed September 17, 2025)
Official websites use .gov
A .gov website belongs to an official government organization in the United States.
Secure .gov websites use HTTPS
A lock (
) or https:// means you’ve safely connected to the .gov website. Share sensitive information only on official, secure websites.
Topological Initialization for Multidimensional Scaling
Published
Author(s)
,
Abstract
Multidimensional scaling is a popular technique for visualizing dissimilarities between objects in complex datasets. For a single dataset, different initial configurations for multidimensional scaling may produce representations with qualitatively different features. This presents both challenges and opportunities for those who use these methods. We introduce the new technique Topological Initialization for Multidimensional Scaling (TIMDS) which employs cycle representatives, a tool from topological data analysis, to generate multidimensional scaling initializations. We show that for some datasets, TIMDS produces representations with competitive stress values and better visualization of key dataset attributes compared to random or classical initialization methods.Citation
Technical Note (NIST TN) - 2349
Report Number
2349
NIST Pub Series
Pub Type
NIST Pubs
Download Paper
Keywords
Cycle representatives, multidimensional scaling, topological data analysis.
Citation
Issues
If you have any questions about this publication or are having problems accessing it, please contact [email protected].
Created September 16, 2025