Machine learning algorithms for early diagnosis of diabetes mellitus: A comparative study

Abstract

A huge majority of people all over the globe are coping with the devastating effects of diabetes, and many of them are not being identified early enough. Diabetes has become one of the most common diseases for cause of death, it develops whenever the person's body becomes unable to create sufficient insulin and the level of glucose is increased in the blood. Overall, Diabetes' global influence has exploded in current years and is expected to continue to do so in the near future. Diabetes affects around 463 million people worldwide; By 2045, this figure will have risen to 700 million. In many nations, the number of people with type 2 diabetes is increasing. Diabetes has become one of the most common diseases for the cause of death, leading to major health issues such as blindness, kidney disease, strokes,heart disease, etc. In this research, some Machine Learning Algorithms are utilized in the prediction ofDiabetes because they are effective at aiding and making predictions from big amounts of data. This research conducts a comparativeevaluation of diabetes diagnosis based on Machine LearningAlgorithms like Naïve Bayes (NB), Support Vector Machine (SVM), Neural Network, Adaboost, K-nearest neighbor (KNN), Linear Kernel SVM etc. In a result, Neural Network produced the best results with the highest accuracy rate.