Labelling strategies for hierarchical multi-label classification techniques
Triguero, Isaac and Vens, Celine (2016) Labelling strategies for hierarchical multi-label classification techniques. Pattern Recognition, 56 . pp. 170-183. ISSN 0031-3203
Many hierarchical multi-label classiﬁcation systems predict a real valued score for every (instance, class) couple, with a higher score reﬂecting more conﬁdence that the instance belongs to that class. These classiﬁers leave the conversion of these scores to an actual label set to the user, who applies a cut-oﬀ value to the scores. The predictive performance of these classiﬁers is usually evaluated using threshold independent measures like precision-recall curves. However, several applications require actual label sets, and thus an automatic labelling strategy.
Actions (Archive Staff Only)