Diabetes predictions have gained major attention due to its consequences on the healthy well-being of an individual. When glucose levels go high due to non- availability of the hormone called insulin which digest glucose, together with other side effects like frequent urination, excessive thirst, and hunger with sudden weight reduction, one can be confirmed of suffering from diabetes. This requires a consistent treatment and monitoring of its complications which are considered fatal in some cases. There are various ways to keep a tract of the glucose level in blood to adjust the diet and dosage of insulin. However, predicting it as early as possible is a challenging task due to its inter-dependency factor that causes trouble to human organs like viscera, peripherals, nervous system, cardiovascular, eyes and excretory system. This research paper aims to provide five different machine learning methods for the prediction of diabetes such as SVM, Logistics regression, KNN Classifier, Random Forest and Logistic algorithm. These proposed methods are effective techniques for earlier detection of the diabetes.