 |
Portfolio data preparation
- Automatic and manual sampling
- Good/Bad records marking
- Visual "Good" and "Bad" borrowers representation
- Graphical statistics
- Easy column and records inclusion/exclusion
- Portfolio data filtering, sorting and re-assignment
- Separate statistics for training and out-of-sample validation subsets
- Portfolio data import from MS Excel, CSV and SAS data files
|
|
|
|
 |
Variables selection
- Goods/Bads distributions
- Bad Rate and WOE for each category
- Logical Trend - Easy analysis of WOE
- Single-click automatic binning with any of 3 algorithms
- Simple and flexible manual binning with drag-n-drop
- Binning and distribution visualization
- Information values for each variable
- Separate statistics for training and out-of-sample validation subsets
|
|
|
|
 |
Scorecard design
- Forward stepwise and backward stepwise regression
- Auto adjustment for prior probabilities
- Auto scaling by odds
- Score points range selection
- Category encoding method selection (1-of-N or WOE)
- Scoring points for every category can be manually changed
- Visual cut-off point definition
- Visual score bands definition
|
|
|
|
 |
Graphs and Reports
- ROC curve
- K-S curve
- Strategy curve
- Marginal event vs Score
- Probability Distribution
- Score Points Distribution
- Risk Segments
- Approval vs Bad Rate
- Chi-squared test with intuitive visualization
- Export to MS Excel
|
|
|
|
 |
Reject Inference (RI)
- Load rejects from CSV or MS Excel file
- Bad Rate comparison for accepts and rejects
- RI is done using the Fuzzy Augmentation method
- Comparison of scorecards before and after RI
- RI can be reverted with a single mouse click
|
|
|
|
