Overview:
This three-day course is intended to equip participants with the skills they need to evaluate the quality of in-house and third-party statistical scorecards and find ways to improve them. The course contains descriptions of the major bugs and bottlenecks in the scorecard development process; the course also gives participants insights into the reject inference approach and the methods used for scorecard development with scarce data.
Duration: 24 hours
Who should attend?
Risk managers, Risk analyst, Predictive analysts from financial, insurance, retail or telecom organizations.
Learn how to:
- Understand the difference between in-house and third-party statistical scorecards
- Generate and analyze scorecard performance graphs and reports
- Monitor (credit scoring) scorecard performance using front-end and back-end reports, PSI and The Chi-Square approach
- Develop representative data samples
- Work with small data samples
- Work with low-characteristics portfolios
- Choose optimal scorecard development techniques
- Improve scorecard quality with binning (classing) techniques
- Improve scorecard quality with reject inference techniques
- Choose optimal approaches for data anomaly transformation
All our scoring trainings.