Nonlinear dimensionality reduction for clustering
dc.creator | Tasoulis S., Pavlidis N.G., Roos T. | en |
dc.date.accessioned | 2023-01-31T10:06:52Z | |
dc.date.available | 2023-01-31T10:06:52Z | |
dc.date.issued | 2020 | |
dc.identifier | 10.1016/j.patcog.2020.107508 | |
dc.identifier.issn | 00313203 | |
dc.identifier.uri | http://hdl.handle.net/11615/79630 | |
dc.description.abstract | We introduce an approach to divisive hierarchical clustering that is capable of identifying clusters in nonlinear manifolds. This approach uses the isometric mapping (Isomap) to recursively embed (subsets of) the data in one dimension, and then performs a binary partition designed to avoid the splitting of clusters. We provide a theoretical analysis of the conditions under which contiguous and high-density clusters in the original space are guaranteed to be separable in the one-dimensional embedding. To the best of our knowledge there is little prior work that studies this problem. Extensive experiments on simulated and real data sets show that hierarchical divisive clustering algorithms derived from this approach are effective. © 2020 | en |
dc.language.iso | en | en |
dc.source | Pattern Recognition | en |
dc.source.uri | https://www.scopus.com/inward/record.uri?eid=2-s2.0-85086990376&doi=10.1016%2fj.patcog.2020.107508&partnerID=40&md5=3c5cf48cd49bfcda5a16cbe2336e4baa | |
dc.subject | Dimensionality reduction | en |
dc.subject | Hierarchical clustering | en |
dc.subject | Divisive hierarchical clustering | en |
dc.subject | High density clusters | en |
dc.subject | Isometric mapping | en |
dc.subject | Nonlinear dimensionality reduction | en |
dc.subject | Nonlinear manifolds | en |
dc.subject | One dimension | en |
dc.subject | Real data sets | en |
dc.subject | Clustering algorithms | en |
dc.subject | Elsevier Ltd | en |
dc.title | Nonlinear dimensionality reduction for clustering | en |
dc.type | journalArticle | en |
Fichier(s) constituant ce document
Fichiers | Taille | Format | Vue |
---|---|---|---|
Il n'y a pas de fichiers associés à ce document. |