BUILDING A MULTIPLE LOGISTIC REGRESSION MODEL TO IDENTIFY FACTORS INFLUENCING AUTISM PATIENTS IN DHI QAR GOVERNORATE

Abstract

The most commonly used statistical method in the field of categorical data analysis is logistic regression, which can be used as a linear model in the field of data classification without assuming the conditions for the distribution of independent variables. The meaning and importance of the description lies in the relationship between the dependent variable and the explanatory variable, and this study focuses on the multi-responsive logistic regression model and the process of formulating the probability of the logistic regression model after estimating its coefficients. The method of maximum probability, when the dependent variable follows the Bernoulli distribution, is the probability of taking the value 1 is p, and the probability of taking the value (0) is 1-p, called binary logistic regression, i.e. response and non-response, in our research was used a multi-responsive logistic regression model, used for dependent variables is nominal and consists of more than two levels, where the basic concepts of binary and multiple logistic regression were touched upon and touched on the most important methods of estimation its features and touched on the methods of testing the model.