Dynamic Score Combination: A Supervised and Unsupervised Score Combination Method
| Title | Dynamic Score Combination: A Supervised and Unsupervised Score Combination Method |
| Publication Type | Conference Paper |
| Year of Publication | 2009 |
| Authors | Tronci, R, Giacinto, G, Roli, F |
| Editor | Perner, P |
| Conference Name | Machine Learning and Data Mining in Pattern Recognition (MLDM 2009) |
| Volume | 5632 |
| Pagination | 163-177 |
| Publisher | Springer |
| Conference Location | Leipzig, Germany |
| Keywords | AUC, bio00, bio02, mcs00, mcs01, ROC, score combination |
| Abstract | In two-class score-based problems the combination of scores from an ensemble of experts is generally used to obtain distributions for positive and negative patterns that exhibit a larger degree of separation than those of the scores to be combined. Typically, combination is carried out by a "static" linear combination of scores, where the weights are computed by maximising a performance function. These weights are equal for all the patterns, as they are assigned to each of the expert to be combined. In this paper we propose a "dynamic" formulation where the weights are computed individually for each pattern. Reported results on a biometric dataset show the effectiveness of the proposed combination methodology with respect to "static" linear combinations and trained combination rules.
|
| Notes | |
| Citation Key | 773 |
