Abstract: Thesystem allows the user to make a use of algorithms to predict the risk ofdiabetes mellitus in human body. The various classification models such asDecision Tree, artificial Neural Networks, Logistic Regression, Associationrules and Naive Bayes are used in this system. Then the Random Forest techniqueis used to find the accuracy of each model in the project. The dataset used isthe Pima Indians Diabetes Data Set, which has the information of patients, someof them have developing diabetes therefore, this project is aimed to create amobile application for predicting persons class whether present in of thediabetes risk or not.

Keywords: DiabetesMellitus, Diabetes Prediction, International Diabetes Federa tion, ElectronicMedical Records