Özet:

f  assessing the Gini indicator, the Kolmogorov–Smirnov statistics and the Mahalanobis distance using the Data Manipulation Language (DML) possibilities within the Structured Query Language (SQL) as the  Fourth-Generation  programming  Language (4GL) implementation providing the corresponding program code was developed. The key feature of the program implementation is the application of the common table expressions, aggregate and analytic windowing functions, table joins, set operations and other possibilities of the DML language within the SQL language as the 4GL approach using the database application Oracle Database 11g as an example. The way of assessing the forecasting performance indicators for an abstract fuzzy probabilistic classifier, particularly in credit scoring, is proposed. The research results are the formalization of the way of assessing the scorecard performance key statistical indicators and providing corresponding program code with the SQL language. The advantages of the proposed way of assessing the statistical indicators with the Fourth-Generation programming Languages are given.

Anahtar Kelimeler: