Scalable multi-output label prediction: from classifier chains to classifier trellises

DOI10.1016/J.PATCOG.2015.01.004MaRDI QIDQ1678708zbMATH OpenOpenAlexDBLPWikidataFDO

Authors Jesse Read, Luca Martino, Pablo M. Olmos, David Luengo

Publication date 20 November 2017

Published in Pattern Recognition (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1501.04870

zbMATH Keywords

Bayesian networksmulti-label classificationclassifier chainsmulti-output predictionstructured inference

Mathematics Subject Classification ID

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

Abstract: Multi-output inference tasks, such as multi-label classification, have become increasingly important in recent years. A popular method for multi-label classification is classifier chains, in which the predictions of individual classifiers are cascaded along a chain, thus taking into account inter-label dependencies and improving the overall performance. Several varieties of classifier chain methods have been introduced, and many of them perform very competitively across a wide range of benchmark datasets. However, scalability limitations become apparent on larger datasets when modeling a fully-cascaded chain. In particular, the methods' strategies for discovering and modeling a good chain structure constitutes a mayor computational bottleneck. In this paper, we present the classifier trellis (CT) method for scalable multi-label classification. We compare CT with several recently proposed classifier chain methods to show that it occupies an important niche: it is highly competitive on standard multi-label problems, yet it can also scale up to thousands or even tens of thousands of labels.

Recommendations

• An analysis of chaining in multi-label classification
• Classifier chains: a review and perspectives
• Efficient Monte Carlo methods for multi-dimensional learning with classifier chains
• Decision functions for chain classifiers based on Bayesian networks for multi-label classification
• Expressive power of binary relevance and chain classifiers based on Bayesian networks for multi-label classification

Cites work

• Bayesian network structure learning by recursive autonomy identification
• Bayesian reasoning and machine learning.
• Classification by pairwise coupling
• Efficient Monte Carlo methods for multi-dimensional learning with classifier chains
• Estimating high-dimensional directed acyclic graphs with the PC-algorithm
• Multilabel classification through random graph ensembles
• On label dependence and loss minimization in multi-label classification
• Statistical comparisons of classifiers over multiple data sets

Cited in

(7)

• Efficient Monte Carlo methods for multi-dimensional learning with classifier chains
• Exact and efficient top-\(K\) inference for multi-target prediction by querying separable linear relational models
• An analysis of chaining in multi-label classification
• Progressive random \(k\)-labelsets for cost-sensitive multi-label classification
• Classifier chains: a review and perspectives
• Using tensor products to detect unconditional label dependence in multilabel classifications
• Decision functions for chain classifiers based on Bayesian networks for multi-label classification

Describes a project that uses

Uses Software

• RAIcode
• pcalg

This page was built for publication: Scalable multi-output label prediction: from classifier chains to classifier trellises

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1678708)