Change search
ReferencesLink to record
Permanent link

Direct link
Simultaneous Predictive Gaussian Classifiers
KTH, School of Engineering Sciences (SCI), Mathematics (Dept.), Mathematical Statistics.
2016 (English)In: Journal of Classification, ISSN 0176-4268, E-ISSN 1432-1343, 1-30 p.Article in journal (Refereed) PublishedText
Abstract [en]

Gaussian distribution has for several decades been ubiquitous in the theory and practice of statistical classification. Despite the early proposals motivating the use of predictive inference to design a classifier, this approach has gained relatively little attention apart from certain specific applications, such as speech recognition where its optimality has been widely acknowledged. Here we examine statistical properties of different inductive classification rules under a generic Gaussian model and demonstrate the optimality of considering simultaneous classification of multiple samples under an attractive loss function. It is shown that the simpler independent classification of samples leads asymptotically to the same optimal rule as the simultaneous classifier when the amount of training data increases, if the dimensionality of the feature space is bounded in an appropriate manner. Numerical investigations suggest that the simultaneous predictive classifier can lead to higher classification accuracy than the independent rule in the low-dimensional case, whereas the simultaneous approach suffers more from noise when the dimensionality increases.

Place, publisher, year, edition, pages
Springer-Verlag New York, 2016. 1-30 p.
Keyword [en]
Bayesian modeling, Discriminant analysis, Inductive learning, Predictive inference, Probabilistic classification
National Category
Probability Theory and Statistics
URN: urn:nbn:se:kth:diva-188231DOI: 10.1007/s00357-016-9197-3ScopusID: 2-s2.0-84959159530OAI: diva2:935922

QC 20160613

Available from: 2016-06-13 Created: 2016-06-09 Last updated: 2016-06-13Bibliographically approved

Open Access in DiVA

No full text

Other links

Publisher's full textScopus

Search in DiVA

By author/editor
Koski, Timo
By organisation
Mathematical Statistics
In the same journal
Journal of Classification
Probability Theory and Statistics

Search outside of DiVA

GoogleGoogle Scholar
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: 5 hits
ReferencesLink to record
Permanent link

Direct link