0.25
0.5
0.75
1.25
1.5
1.75
2
Local Decomposition for Rare Class Analysis
Published on Aug 14, 20073151 Views
Given its importance, the problem of predicting rare classes in large-scale multi-labeled data sets has attracted great attentions in the literature. However, the rare-class problem remains a critical
Chapter list
Local Decomposition for Rare Class Analysis00:03
Outline pt 100:26
Rare Class Analysis00:47
Research Motivation – Problems01:36
Problem Formulation02:41
Our Contributions03:07
Outline pt 203:41
Directions and Objectives of Our Method03:49
Algorithm Description05:09
An Example06:16
Effect of COG&COG-OS on Rare Class pt 107:24
Effect of COG&COG-OS on Rare Class pt 208:05
Outline pt 308:56
Experimental Design09:02
The Experimental Setup09:22
1-1: COG on Imbalanced 2-class Data pt 109:58
1-1: COG on Imbalanced 2-Class Data pt 210:56
1-2: COG on Imbalanced Multi-Class Data11:32
1-3: COG vs. Resampling 12:01
1-4: COG on KDDCUP99 Data12:51
2-1: COG on Balanced Data pt 113:51
2-1: COG on Balanced Data pt 214:41
2-2: COG vs. Random Partitioning15:05
3: Discussion on Feature Selection15:44
Outline pt 416:48
Related Work16:54
Outline pt 517:31
Concluding Remarks17:35
Thank You!18:07