DIABETES PREDICTION USING MACHINE LEARNING

Diabetes is a chronic disease which occurs when the level of glucose rises above a certain amount. In other words, this is the case when the pancreas stops producing the necessary amount of insulin which controls the level of blood glucose. According to International Diabetes Federation, almost 382 million people are living with diabetes across the whole world. By 2035, the number of people with diabetes is forecast to increase up to 592 million. Diabetes is considered to be the major cause of blindness, stroke, kidney failure and many other fatal illnesses. When we consume food, our blood turns the food into sugar, or glucose. At that point, our pancreas normally releases a hormone called insulin. Insulin allows the glucose from a person’s food to access the cells in their body to supply energy. However, when a person has diabetes, this system does not work. It is generally known that Type 1 and Type 2 are the most common forms of diabetes among the elderly as well as adults, but there are also other forms such as gestational diabetes which occurs during pregnancy, and others.

Machine learning is becoming a leading scientific filed in data science dealing with the ways in which machines themselves learn from experience and gradually develop. The purpose of this project is to develop a system which enables to diagnose diabetes in patients at early stage so that doctors can prevent serious consequences effectively. There are various algorithms such as K nearest neighbor, Logistic Regression, Decision Tree, Super Vector Machine and others to predict illness with high accuracy. In this project, we have decided to use Super Vector Machine algorithm in order to predict diabetes based on several symptoms in patients, and we have used a dataset which involves the symptoms of diabetes.