DIABETES RISK ASSESSMENT USING MACHINE LEARNING: A COMPARATIVE STUDY OF CLASSIFICATION ALGORITHMS

Abstract

Diabetes is a serious health condition with high blood glucose/sugar levels. Diabetes is a chronicle disease that can cause worldwide health care crisis but we can take some steps to manage these crisis. IN Diabetes Blood sugar/glucose is the main source of energy that is drawn from the food we eat in our day to day life.

Insulin is a hormone that is produced by the pancreases in our body which helps the glucose gets into the cells which can be used for energy to perform day to day activities. When body doesn’t make enough or any insulin then glucose stays in the blood which might lead to various health problems like heart attack, strokes etc.

There are many types of diabetes like type1, type2, gestational and monogenic diabetes where type1 and type2 are the most common ones.type1 is mostly diagnosed in young adults and children and type2 is mostly diagnosed in middle-age or older group of people.

Machine learning is a scientific field here machine learn from the human experiences the aim of the project is to build a system which can predict whether the patient is diabetic or not with a high accuracy by combining result of various machine learning technique with the algorithm used like KNN, logistic regression, random forest etc.

The accuracy of the model using each algorithm is calculated then the one with more percentage of accuracy is taken as the model for predicting diabetes.