Comparison of Machine Learning Classification Algorithms for Prediction of Early-stage Diabetes

Shirish Mohan  Dubey; Vani  Chaturvedi; Samidha  Bafna; Sumit Sing  Rathore; Rajat  Tank

Abstract

Diabetes is the most widespread and gruesome disease spread all around the globe. It is a chronic metabolic disease which has mild symptoms which are hard to spot at the initial stage. It is well known that diabetes causes a high blood sugar level. This is because insulin is not able to carry sugar from your blood into your cells in order to be stored or used for energy as it should. Negligence in treatment and identification of this disease can damage a person’s nerves, eyes, kidneys, skin, and other organs. Thus, timely diagnosis and treatment can be effective in overcoming this disease. The purpose of this study is to review a number of frameworks that have been proposed as ways to predict diabetes disease and classify diabetes using Machine Learning (ML) algorithms. The database is a summarized set of data collected from various medical research centres and hospitals from across the India. The emergence of diabetes can be predicted by applying four different classification algorithms, i.e., KNN Algorithm, Logistic Regression, Random Forest, and Support Vector Machine. The motive of this framework is to save the money and time of a patient in diagnosis using various ML approaches.

Abstract

Keywords