Change search
ReferencesLink to record
Permanent link

Direct link
Bayesian unsupervised classification framework based on stochastic partitions of data and a parallel search strategy
University of Helsinki. (Department of mathematics and statistics)
University of Helsinki. (Department of Mathematics)
KTH, School of Engineering Sciences (SCI), Mathematics (Dept.), Mathematical Statistics. (computational biostatistics)ORCID iD: 0000-0003-1489-8512
2009 (English)In: Advances in Data Analysis and Classification, ISSN 1862-5347, Vol. 3, no 1, 3-24 p.Article in journal (Refereed) Published
Abstract [en]

Advantages of statistical model-based unsupervised classification over heuristic alternatives have been widely demonstrated in the scientific literature. However, the existing model-based approaches are often both conceptually and numerically instable for large and complex data sets. Here we consider a Bayesian model-based method for unsupervised classification of discrete valued vectors, that has certain advantages over standard solutions based on latent class models. Our theoretical formulation defines a posterior probability measure on the space of classification solutions corresponding to stochastic partitions of observed data. To efficiently explore the classification space we use a parallel search strategy based on non-reversible stochastic processes. A decision-theoretic approach is utilized to formalize the inferential process in the context of unsupervised classification. Both real and simulated data sets are used for the illustration of the discussed methods.

Place, publisher, year, edition, pages
Springer Berlin/Heidelberg, 2009. Vol. 3, no 1, 3-24 p.
Keyword [en]
Bayesian classification, Markov chain Monte Carlo, Statistical learning, Stochastic optimization
National Category
Computer and Information Science
URN: urn:nbn:se:kth:diva-83214DOI: 10.1007/s11634-009-0036-9ScopusID: 2-s2.0-67651049356OAI: diva2:498790
QC 20120214Available from: 2012-02-14 Created: 2012-02-12 Last updated: 2012-02-14Bibliographically approved

Open Access in DiVA

parallell.pdf(547 kB)605 downloads
File information
File name FULLTEXT01.pdfFile size 547 kBChecksum SHA-512
Type fulltextMimetype application/pdf

Other links

Publisher's full

Search in DiVA

By author/editor
Koski, Timo
By organisation
Mathematical Statistics
In the same journal
Advances in Data Analysis and Classification
Computer and Information Science

Search outside of DiVA

GoogleGoogle Scholar
Total: 605 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

Altmetric score

Total: 39 hits
ReferencesLink to record
Permanent link

Direct link