Predicting Diabetes Mellitus Using Feature Selection And Classification Techniques In Machine Learning Algorithms

Diabetes is a disease that is now spreading like an epidemic around the globe. Diabetics is a chronic disease that occurs when the blood sugar or glucose in the body is not controlled or broken down. It may be caused either when the body does not react to the insulin produced naturally in the body or when the produced insulin is insufficient. The latest WHO statistics points diabetics as a life-threatening disease condition with an estimated 1.6 million deaths worldwide. The word diabetics mellitus is of Greek origin that means 'to pass through honey or sweet'. Constant high blood sugar in blood stream termed hyperglycemia is a serious condition that can adversely affect the health of an individual. A patient may experience loss of energy with fatigue and brokenness. Uncontrolled levels threaten body organs which include kidneys, heart, eyes and nervous system. Taking into account the widespread nature of the disease, finding a cure using latest computer advancements has been a topic of study for many researchers and scientists worldwide. This research focuses on creating a forecast or a prediction algorithm that can sort out an optimal classifier. The optimal classifier must be able to deliver near close results to real world clinical outcomes when it is juxtaposed to a validity of its accuracy. Sorting out attributes that trouble early detection of the disease is the objective of the study.