Development of a binary logistic regression model for individuals with rheumatic diseases

Ouraim in this study is tocreate a logistic regression model for individuals witht rheumatic disease using the least possible explanatory variables with a high classification rate. Thus, the physician specializing in the sediseases will be able to predict the presence of rheumatic disease at the lowest cost and witht the least effort,, using a logistic regression model that will give the probability and percentage of the disease. Thus, the doctor can select the cases that need carefulexamination and perform the necessary laboratory tests. Since the setestsare very expensive and have an important place in affecting the economy of many countries, usingthis model will bothaccelerate the process and provide economic benefits by ensuring that it is applied to the patients who need it. Also, in this the sis, we will discuss the various problem sthatthe logistic model will fail and how tosolvethe model. We will address two serious problem sthat completelydestroy the logistic model, the singularity problem and the multicollinearity problem. Thus, we will make an important contribution to the literature and scientists working on the seissues. Keywords: Logistic Regression, Rheumatic Disease