Efficient Voting Prediction for Pairwise Multilabel Classification
Type of publication:  Inproceedings 
Citation:  jf:ESANN09 
Booktitle:  Proceedings of the 17th European Symposium on Artificial Neural Networks (ESANN 2009, Bruges, Belgium) 
Year:  2009 
Month:  April 
Pages:  117122 
Publisher:  dside publications 
ISBN:  2930307099 
URL:  http://www.dice.ucl.ac.be/Proceedings/esann/esannpdf/es2009112.pdf 
Abstract:  The pairwise approach to multilabel classification reduces the problem to learning and aggregating preference predictions among the possible labels. A key problem is the need to query a quadratic number of preferences for making a prediction. To solve this problem, we extend the recently proposed QWeighted algorithm for efficient pairwise multiclass voting to the multilabel setting, and evaluate the adapted algorithm on several realworld datasets. We achieve an averagecase reduction of classifier evaluations from n^2 to n + dn log n, where n is the total number of labels and d is the average number of labels, which is typically quite small in realworld datasets. 
Userfields:  opturl={"http://www.ke.tudarmstadt.de/publications/papers/ESANN09.pdf"}, 
Keywords:  
Authors  
Topics


