An empirical study of applying ensembles of heterogeneous classifiers on imperfect data
PAKDD'09 Proceedings of the 13th Pacific-Asia international conference on Knowledge discovery and data mining: new frontiers in applied data mining
Borderline over-sampling for imbalanced data classification
International Journal of Knowledge Engineering and Soft Data Paradigms
Building decision trees for the multi-class imbalance problem
PAKDD'12 Proceedings of the 16th Pacific-Asia conference on Advances in Knowledge Discovery and Data Mining - Volume Part I
| Hi-index | 0.00 |
Class imbalance is a ubiquitous problem in supervised learning and has gained wide-scale attention in the literature. Perhaps the most prevalent solution is to applysampling to training data in order improve classifier performance. The typical approach will apply uniform levels of sampling globally. However, we believe that datais typically multi-modal, which suggests sampling shouldbe treated locally rather than globally. It is the purposeof this paper to propose a framework which first identifies meaningful regions of data and then proceeds to findoptimal sampling levels within each. This paper demonstrates that a global classifier trained on data locally sampled produces superior rank-orderings on a wide range ofreal-world and artificial datasets as compared to contemporary global sampling methods.