Loading...
Loading...

Go to the content (press return)

Goodness-of-fit filtering in classical metric multidimensional scaling with large datasets

Author
Graffelman, J.
Type of activity
Journal article
Journal
Journal of applied statistics
Date of publication
2019-12-18
DOI
10.1080/02664763.2019.1702929
Project funding
Theoretical Population Genetics
Transferring compositional data methods into applied science and technology
Repository
http://hdl.handle.net/2117/175485 Open in new window
URL
https://www.tandfonline.com/doi/abs/10.1080/02664763.2019.1702929 Open in new window
Abstract
Metric multidimensional scaling (MDS) is a widely used multivariate method with applications in almost all scientific disciplines. Eigenvalues obtained in the analysis are usually reported in order to calculate the overall goodness-of-fit of the distance matrix. In this paper, we refine MDS goodness-of-fit calculations, proposing additional point and pairwise goodness-of-fit statistics that can be used to filter poorly represented observations in MDS maps. The proposed statistics are especially ...
Citation
Graffelman, J. Goodness-of-fit filtering in classical metric multidimensional scaling with large datasets. "Journal of applied statistics", 18 Desembre 2019.
Keywords
Allele sharing distance, Attractor point, Eigenvalue, Manhattan distance, Outlier, Plot brushing
Group of research
COSDA-UPC - COmpositional and Spatial Data Analysis

Participants