Enhancing Diabetes Prediction with Data Preprocessing and various Machine Learning Algorithms

Abstract

  Diabetes mellitus, usually called diabetes, is a serious public health issue that is spreading like an epidemic around the world. It is a condition that results in elevated glucose levels in the blood. India is often referred to as the 'Diabetes Capital of the World', due to the country's 17% share of the global diabetes population. It is estimated that 77 million Indians over the age of 18 have diabetes (i.e., everyone in eleven) and there are also an estimated 25 million pre-diabetics. One of the solutions to control diabetes growth is to detect it at an early stage which can lead to improved treatment. So, in this project, we are using a few machine learning algorithms like SVM, Decision Tree Classifier, Random Forest, KNN, Linear regression, Logistic regression, Naive Bayes to effectively predict the diabetes. Pima Indians Diabetes Database has been used in this project. According to the experimental findings, Random Forest produced an accuracy of 91.10% which is higher among the different algorithms used.