Performance Analysis of Supervised Machine Learning Algorithm for Prediction of Diabetes

Diabetes is affecting a lot of people nowadays. It is a very common disease throughout the world. People found it difficult to predict the probability of this disease at an earlier stage. The disease is not curable yet by medical science but it can only be controlled. Furthermore, patients of this disease have to take medicines throughout their life. This disease can be predicted using different statistical tools. This paper proposed a machine learning based model to predict this disease on the basis of medical symptoms of a patient. Three best known machine learning algorithms are applied to predict diabetes which are KNN, Logistic Regression and Random Forest. Models of these machine learning algorithms are implemented using python programming language. The prediction accuracy of these three machine learning models is calculated and compared. The out coming results shows that the Random Forest performs best among these three machine learning models.