On the expected probability of correct classification

Kokolakis, GE

dc.contributor.author	Kokolakis, GE	en
dc.date.accessioned	2014-03-01T01:05:57Z
dc.date.available	2014-03-01T01:05:57Z
dc.date.issued	1981	en
dc.identifier.issn	0006-3444	en
dc.identifier.uri	https://dspace.lib.ntua.gr/xmlui/handle/123456789/9086
dc.subject.classification	Biology	en
dc.subject.classification	Mathematical & Computational Biology	en
dc.subject.classification	Statistics & Probability	en
dc.title	On the expected probability of correct classification	en
heal.type	journalArticle	en
heal.identifier.primary	10.1093/biomet/68.2.477	en
heal.identifier.secondary	http://dx.doi.org/10.1093/biomet/68.2.477	en
heal.language	English	en
heal.publicationDate	1981	en
heal.abstract	SUMMARY: The Bayesian approach is applied to examine how the number of features used in a classification problem affects the performance. Exact and asymptotic results are derived for two particular classification problems, demonstrating the positive effect of increasing the number of features. © 1981 Biometrika Trust.	en
heal.publisher	BIOMETRIKA TRUST	en
heal.journalName	Biometrika	en
dc.identifier.doi	10.1093/biomet/68.2.477	en
dc.identifier.isi	ISI:A1981LZ46500018	en
dc.identifier.volume	68	en
dc.identifier.issue	2	en
dc.identifier.spage	477	en
dc.identifier.epage	483	en