Exploiting label dependencies for improved sample complexity

DOI10.1007/S10994-012-5312-9MaRDI QIDQ374170zbMATHOpenAlexFDO

Authors Lena Chekina, Dan Gutfreund, Lior Rokach, Bracha Shapira, Aryeh Kontorovich

Publication date 22 October 2013

Published in Machine Learning (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1007/s10994-012-5312-9

zbMATH Keywords

multi-label classificationartificial datasetsconditional and unconditional label dependenceempirical experimentensemble learning algorithmsensemble models diversitygeneralization boundsmulti-label evaluation measures

Mathematics Subject Classification ID

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

Recommendations

• On label dependence and loss minimization in multi-label classification
• A Two-Stage Conditional Random Field Model Based Framework for Multi-Label Classification
• Reduced-rank multi-label classification
• Combining instance-based learning and logistic regression for multilabel classification
• Multilabel classifiers with a probabilistic thresholding strategy

Cites work

• scientific article; zbMATH DE number 823069 (Why is no real title available?)
• BoosTexter: A boosting-based system for text categorization
• Convergence of stochastic processes
• Feature selection for multi-label naive Bayes classification
• Genetic algorithm-based feature set partitioning for classification problems
• Learnability and the Vapnik-Chervonenkis dimension
• On label dependence and loss minimization in multi-label classification
• On the Uniform Convergence of Relative Frequencies of Events to Their Probabilities
• Pattern classification using ensemble methods.
• Statistical comparisons of classifiers over multiple data sets
• The VC dimension of \(k\)-fold union
• Toward efficient agnostic learning
• Wrappers for feature subset selection
• \(k\)-Fold unions of low-dimensional concept classes

Cited in

(3)

• A stacking weighted \(k\)-nearest neighbour with thresholding
• Online multi-label dependency topic models for text classification
• Compressed labeling on distilled labelsets for multi-label learning

Describes a project that uses

Uses Software

• MULAN
• BoosTexter

This page was built for publication: Exploiting label dependencies for improved sample complexity

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q374170)