Bayes-Optimal Hierarchical Multilabel Classification
Authors:
- 4 June 2015
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Knowledge and Data Engineering
- Vol. 27 (11) , 2907-2918
Abstract
Hierarchical multilabel classification allows a sample to belong to multiple class labels residing on a hierarchy, which can be a tree or directed acyclic graph (DAG). However, popular hierarchical loss functions, such as the H-loss, can only be defined on tree hierarchies (but not on DAGs), and may also under- or over-penalize misclassifications near the bottom of the hierarchy. Besides, it has been relatively unexplored on how to make use of the loss functions in hierarchical multilabel classification. To overcome these deficiencies, we first propose hierarchical extensions of the Hamming loss and ranking loss which take the mistake at every node of the label hierarchy into consideration. Then, we first train a general learning model, which is independent of the loss function. Next, using Bayesian decision theory, we develop Bayes-optimal predictions that minimize the corresponding risks with the trained model. Computationally, instead of requiring an exhaustive summation and search for the optimal multilabel, the resultant optimization problem can be efficiently solved by a greedy algorithm. Experimental results on a number of real-world data sets show that the proposed Bayes-optimal classifier outperforms state-of-the-art methods.Keywords
Funding Information
- Council of the Hong Kong Special Administrative Region (614012)
This publication has 25 references indexed in Scilit:
- Generalized k-Labelsets Ensemble for Multi-Label and Cost-Sensitive ClassificationIEEE Transactions on Knowledge and Data Engineering, 2013
- Hierarchical Multilabel Classification with Minimum Bayes Risk2012
- Compressed labeling on distilled labelsets for multi-label learningMachine Learning, 2012
- Transductive Multilabel Learning via Label Set PropagationIEEE Transactions on Knowledge and Data Engineering, 2011
- Multi-label learning by exploiting label dependency2010
- A categorization approach to automated ontological function annotationProtein Science, 2006
- Hierarchical classification2006
- Automatically learning document taxonomies for hierarchical classification2005
- Hierarchical document categorization with support vector machines2004
- Large margin hierarchical classification2004