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

PATTERN RECOGNITION

卷:48

Article

Read, Jesse^1,2  Martino, Luca³  Olmos, Pablo M.⁴  Luengo, David⁵ 

[1] Aalto Univ, Dept Comp Sci, Helsinki, Finland

[2] Aalto Univ, Helsinki, Finland

[3] Univ Helsinki, Dept Math & Stat, Helsinki, Finland

[4] Univ Carlos III Madrid, Dept Signal Theory & Commun, Madrid 28911, Spain

[5] Univ Politecn Madrid, Dept Signal Theory & Commun, Madrid 28031, Spain

关键词: Classifier chains;    Multi-label classification;    Multi-output prediction;    Structured inference;    Bayesian networks;   

DOI  :  10.1016/j.patcog.2015.01.004

来源: Elsevier

【 摘 要 】

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 modelling a fully cascaded chain. In particular, the methods' strategies for discovering and modelling a good chain structure constitute a major 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. (C) 2015 Elsevier Ltd. All rights reserved.

【 授权许可】

Free