Diabetes is a chronic, gradually progressing metabolic disease that is complex and due to which the patient suffers from elevated glucose levels in the blood. Type1, Type2, and gestational diabetes are the three common diabetes types. It causes severe health damages like foot ulcers, vision impairment, kidney disorders, and cardiac strokes. Hyperglycemia is one of its major manifestations. Identifying the progress of the disease at an early stage helps in the prevention of further complications. Machine learning methods are most frequently used in the classification of Quantitative data. It facilitates in identifying the relevant factors that most contribute to disease development. The least significant attributes are eliminated during the pre-processing phase. In this paper, the classification or prediction of the disease Diabetes Mellitus using traditional supervised machine learning methods was discussed and their prediction accuracy was analyzed. Several tools, datasets, and different metrics are studied and the frequently used efficient machine learning model among the several existing algorithms are investigated. The Deep learning algorithm of Artificial Neural Networks is one of the most efficient Machine learning Models that is preferred by many scientists and researchers in the health domain.

.