Name
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. |
Year | DOI | Venue |
|---|---|---|
2009 | 10.1007/978-3-642-15751-6_37 | CLEF (2) |
Keywords | Field | DocType |
different visual feature,ranking svm,classical svm,imbalanced classification problem,standard svm,convenient loss function,image annotation,chosen loss function,imbalanced data,prediction quality,improving image annotation,loss function | Automatic image annotation,Ranking SVM,Pattern recognition,Computer science,Support vector machine,Artificial intelligence,Classifier (linguistics),Machine learning | Conference |
Volume | ISSN | ISBN |
6242 | 0302-9743 | 3-642-15750-5 |
Citations | PageRank | References |
5 | 0.61 | 3 |
Authors | ||
4 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Ali Fakeri-Tabrizi | 1 | 37 | 6.07 |
| Sabrina Tollari | 2 | 91 | 13.64 |
| Nicolas Usunier | 3 | 1974 | 97.52 |
| Patrick Gallinari | 4 | 1856 | 187.19 |