Change Log¶
Version 0.5 (2016-01-29)¶
- Support of Python 3.5
Version 0.4 (2015-05-26)¶
- Documentation
- Cost-sensitive Decision Trees algorithm
- Cost-sensitive Ensemble methods
- Tutorial CS Credit Scoring
- Slides CS Credit Scoring
Version 0.03 (2014-09-19)¶
- Initial public release
- Include the cost-sensitive algorithms:
- Bayes minimum risk classifier
- Cost-sensitive Logistic Regression
- Cost-sensitive Thresholding
- Sampling techniques:
- Cost-proportionate sampling methods
- SMOTE
- under-sampling
- ROCCH probability calibration
- Evaluation metrics
- Cost-sensitive datasets:
- Direct Marketing
- Kaggle11 Credit Scoring
- PAKDD09 Credit Scoring