RESEARCH ON BUILDING LOGICSTIC REGRESSION MODEL BY R SOFTWARE TO DIABETES EVALUATE AND PREDICTION

Abstract

Problem: In commune and ward hospitals far from the centre, diabetes diagnosis is still not on time because it is difficult to access to test. Objectice: The paper proposes a solution applied Logistic regression model combined with integrated libraries on R software to select the optimal model for diabetes prediction and assessment which will save costs and promptly diagnose for patients in communes and wards far from the centre that cannot afford testing. Setting and method:  The research is performed in Thai Nguyen university of medicine and pharmacy by experiment. The authors uses R software to run Logistic regression model on available diabetes data set. Results: The results show that the model has a very good classification with the sensitivity reaching 80.60% and the specificity reaching 93.26%.