Credit Scoring Steps
- Data - includes loading the data into CreditScoring as
well as selecting correct modelling samples
- Target and Partitioning - during this step the target variable will be selected
and analysed, as well as different subsets of the modelling data set used for
development, validation and testing which will be explained in the next chapter
- Grouping - presents the most crucial modelling step, where measures
such as Weight of Evidence (WoE) and Information Value (IV) are calculated.
In a nutshell they help determine how input variables and their corresponding
attributes affect the target variable.
Grouping of input variables into
distinct categories is also performed in this step.
- Selection - used to determine which variables will be used in the regression process
- Scorecard - helps us determine the final scorecard and creates charts
that visualize corresponding measures that describe the created
scorecard’s ability to correctly classify “default” and “non-default”