1- PhD Student in Accounting, Shahroud Branch, Islamic Azad University, Shahroud, Iran (Corresponding Author) , amiri.hiva@yahoo.com 2- Department of Accounting, Shahroud Branch, Islamic Azad University, Shahroud, Iran 3- Department of Accounting, Shahroud Branch, Islamic Azad University, Shahroud, Iran.
Abstract: (2204 Views)
This article aims to model the credit risk assessment and credit assessment of banks and non-banking institutions by Linear Discriminant Analysis ،logit regression and probit.
For this purpose, the statistical sample size is determined through the "screening sampling method". The researcher collects observations by sampling members of the statistical community by screening sampling method This sample size is included during the financial periods of the years 91 to 98. In this article, after reviewing the financial statements of each bank and non-banking institution, the explanatory variables are evaluated.
In the general comparison between the predictive accuracy of linear Discriminant analysis method and nonlinear methods of logistic regression and probit, the results of this comparison showed that the Linear Discriminant Analysis is on the same level with probit regression analysis in terms of accuracy and efficiency, while Logistic regression analysis was different and less efficient than these two methods.
amiri H, dehdar F, Abdoli M. An optimal model for estimation of Probability of Default related to banks and non banking credit institutions on the basis of Linear Discriminant Analysis and Nonlinear Probability Models of Logit & Probit. mieaoi 2021; 10 (36) : 6 URL: http://mieaoi.ir/article-1-1088-en.html