Improving Image Annotation in Imbalanced Classification Problems with Ranking SVM

Abstract : We try to overcome the imbalanced data set problem in image annotation by choosing a convenient loss function for learning the classifier. Instead of training a standard SVM, we use a Ranking SVM in which the chosen loss function is helpful in the case of imbalanced data. We compare the Ranking SVM to a classical SVM with different visual features. We observe that Ranking SVM always improves the prediction quality, and can perform up to 23% better than the classical SVM.
Liste complète des métadonnées

https://hal.archives-ouvertes.fr/hal-00581661
Contributor : Sabrina Tollari <>
Submitted on : Thursday, March 31, 2011 - 2:43:23 PM
Last modification on : Thursday, March 21, 2019 - 2:41:32 PM

Links full text

Identifiers

Citation

Ali Fakeri-Tabrizi, Sabrina Tollari, Nicolas Usunier, Patrick Gallinari. Improving Image Annotation in Imbalanced Classification Problems with Ranking SVM. CLEF 2009 - 10th Workshop of the Cross-Language Evaluation Forum, Sep 2009, Corfu, Greece. pp.291-294, ⟨10.1007/978-3-642-15751-6_37⟩. ⟨hal-00581661⟩

Share

Metrics

Record views

88