Тип публикации: доклад, тезисы доклада, статья из сборника материалов конференций
Год издания: 2010
Ключевые слова: Eventological scoring, Wide-multiplicative projections and approximations of eventological distributions, Comparative analysis, Credit risks, Probabilistic dependence, Probability measures, Automation, Information technology, Risk assessment, Probability distributions
Аннотация: Eventological scoring (e-scoring) is generalization of classical bank scoring as allows to consider structure of dependencies of base events. E-scoring gives probabilistic vindication using formulas of linear multifactorial and logistic regress, which are true only for two kinds of dependencies of structures of base events, and, besides it allows estimating probability of an objective event for the enclosed structure. Thus, eventological scoring allows to carry out the comparative analysis with a view of diagnostics of possible credit risk. The work purpose is possibility expansion of eventological scoring by means of the theory of wide dependence of events. Wide dependence of events is a community of structures of probabilistic dependence of events which are determined by characteristic relations and delimitation, which must satisfy to probability measure. Using special tools of wide- projections and approximations of eventological distributions, it is possible to define probability of the occurrence of object events in the conditions of any statistics by a method of eventological scoring for any structure of dependencies of base events.
Издание
Журнал: Proceedings of the IASTED International Conference on Automation, Control, and Information Technology - Control, Diagnostics, and Automation, ACIT-CDA 2010
Номера страниц: 316-322
Персоны
- Goldenok E.E.
- Lukyanova N.A.
- Semenova D.V.
- Goldenok E.E. (Krasnoyarsk State Institute of Economics and Trade,Siberian Federal University)
- Lukyanova N.A. (Krasnoyarsk State Institute of Economics and Trade,Siberian Federal University)
- Semenova D.V. (Institute of Mathematics,Siberian Federal University)
Вхождение в базы данных
Информация о публикациях загружается с сайта службы поддержки публикационной активности СФУ. Сообщите, если заметили неточности.