Кластеризация заемщиков - физических лиц по уровню дефолтов: рейтинговый подход (на примере регионов Сибирского федерального округа)
This article suggested method of probability estimation of defaults in the retail. This method is based on the clustering of borrowers depending on the level of credit risk and satisfying international standards of credit risk. Performed testing methodology on the example of some regions of the Siberian Federal District, identified the essential features of the behavior of borrowers in the regions. Calculated the ratings of the regions in terms of probability of defaults on consumer loans (retail).