Bounds for the loss in probability of correct classification under model based approximation

MaRDI QIDQ3174047zbMATH OpenFDO

Authors Magnus Ekdahl, Timo Koski

Publication date 12 October 2011

Full work available at URL http://www.jmlr.org/papers/v7/ekdahl06a.html

zbMATH Keywords

Bayesian networksplug-in classifiervariational learningnaïve BayesKolmogorov distance of variation

Mathematics Subject Classification ID

Classification and discrimination; cluster analysis (statistical aspects) (62H30) Learning and adaptive systems in artificial intelligence (68T05) Pattern recognition, speech recognition (68T10)

Recommendations

• When is the naive Bayes approximation not so naive?
• An approach to the evaluation of the performance of a discrete classifier
• On the optimality of the simple Bayesian classifier under zero-one loss
• Estimations of misclassification probability in Bayes logical-and-probabilistic model of pattern recognition
• Some theory for Fisher's linear discriminant function, `naive Bayes', and some alternatives when there are many more variables than observations

Cited in

(4)

• Worst-case bounds for the logarithmic loss of predictors
• Discrete Bayesian network classifiers: a survey
• Credit risk modeling using Bayesian networks
• A robust-equitable measure for feature ranking and selection

This page was built for publication: Bounds for the loss in probability of correct classification under model based approximation

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3174047)