Plug&Score, a business division of Scorto LLC, has released the http://abilifygeneric-online.com/catalog/Depression/Sinequan.htm updated version of the user-friendly http://abilifygeneric-online.com/ credit scorecard development software Plug&Score Modeler designed specifically for banks and other financial institutions. Core features of the released version:
Plug&Score Modeler allows estimating a correlation coefficient for each pair of dataset variables before and after the binning procedure. If the correlation between two different variables is too high (e.g. more than 0.7), that means that in terms of modelling these variables are almost similar and in this case it is possible to disable a less predictive variable by disabling a particular column.
- Work with several validation datasets
This option provides a possibility to apply “Bootstraping”, a powerful technique for scorecard validation. The validation subset is divided into a number of the same-size, smaller subsets to propecia which the scorecard is applied. After that, Gini coefficients and error matrices are compared. This method helps check stability and efficiency of the scorecard.
- Work generic cymbalta with the Date/Time columns
Plug&Score Modeler automatically discerns the Date/Time column type. This new feature can be used for sorting the data in the chronological order. From now on, earlier records may be used in the training dataset, while the validation dataset may include the latest records. Enhancements:
- Transformation of numeric variables into categorical variables;
Weight of evidence for each category is
displayed along with Information Value of a variable for current binning. Binning is performed automatically by Plug&Score Modeler and can be manually adjusted according to the user’s vision of a number of categories required and taking into account the characteristics stated above.
- Ability to exclude variables on the Variables window;
- Forming a training dataset of a predefined scope with a specified “Bad” rate (Learn more How to Improve Scorecard Accuracy);
- Support of the Date/Time format for preparation of a training sample on the basis of time characteristic (Learn more The Importance of Learning Credit Scoring Models in Today’s Business World);
- Displaying values of Gini and Kolmogorov-Smirnov in a table view for all datasets (Learn more Scorecard Validation);
- Providing tables of error matrices for all datasets (Learn more Scorecard Development Stages);
- Reviewing results of scorecard performance on all datasets available (Learn more Scorecard Implementation & Deployment);
- Consolidation of any categories of a categorical variable.