Data
How does this work in practice? In order to successfully apply regression analysis and determine the relationship between the target variable and independent variables the bank should prepare a data set containing historical data where the values of target and independent variables are already known.
>The data set (collection of data, usually in the form of a single database table) for creating a statistical scorecard should contain historical data about clients to whom loans have already been disbursed to.
Data Structure
Assuming that the data set for modelling has been created from historical data about previously disbursed loans and their corresponding repayment outcome in the following format, record by record:
| ID number | Input Variable 1 | Input Variable 2 | ... | Target variable |
|---|---|---|---|---|
| 1 | Value12 | Value12 | ||
| 2 | Value12 | Value12 |
Supported File Formats
CreditScoring supports the following input file formats:
- Tab delimited (text)
- Comma separated values - CSV (extension .csv)
The software auto-detects the data separator but it also allows for custom separators to be entered by the user.
Importing the Data
CreditScoring software offers the users versatile tools to analyse the data and
to select the most appropriate variables and their parameters easily.
Detailed knowledge of the data set by the user is absolutely crucial
not only for identifying potential data quality errors, but understanding
the structure and distribution of the variables and the underlying business logic.
