Construction features of scoring models based on Deductor analytical platform

The aim of the study is tobuild a scoring model based on Deductor analytical platform.Problems solved by this method are introduced to actualizing theresearch.The source datafor customer response predictionis the scoring profile of the OTP Bank. Modeling processand Fine&Coarse ClassingDeductor's handler are described. Fine&CoarseClassing is capable of reducing the unique values and allows to analyze iterativelyand to improve the quality and speed of model building. The weightofevidence, informationvalueand the value of the objective functionenable the fundamental parameters to be determined from experimental results. The conclusion is made that using Deductor haveseveral advantages over manual calculations.