Investigations of reasons for district disasters

SUMMARY Investıgations of Reasons For District Disasters In this study, between 2005 and 2017 aimed at making a Poisson regression application data on the number of divorces in Turkey. Initially, information on regression analysis is given. Regression is divided into linear regression and non-linear regression. Poisson regression is a nonlinear regression. Poisson regression analysis is used when the dependent variable is a number of data (count). To date, the Poisson regression model has been investigated under the assumption that the mean and variance for the Poisson distribution are equal. In cases of overexploitation, two paths are followed. The first of these is estimating a weighing parameter, and then correcting the test statistics and remnants. The second is negative binomial regression. In the implementation of the divorce data, whether the data in the regions are normally distributed, averages, parameter estimates, whether there are regional differences are examined using the SPSS package program. In practice, divorce data were tried to be modeled with poisson regression. Key Words: Poisson regression, divorce, linear regression, non-linear regression, Negative Binomial Regression