Improving risk predictions by preprocessing imbalanced credit data
ICONIP'12 Proceedings of the 19th international conference on Neural Information Processing - Volume Part II
Hi-index | 0.00 |
This paper performs systematic comparative studies on weighted methods including weight C4.5, weighted SVM and weighted rough set with traditional C4.5, SVM and rough set for credit scoring. The experiments show that the weighted methods outperform to the traditional methods when the methods are sensitive to the class distribution.