Diabetes Prediction using Machine Learning Techniques

Abstract

Now a day due to hectic schedules and sedentary lifestyle people do not follow the proper diet. Poor diet may lead to diabetes, and which could result in various health issues such as heart attacks, strokes, renal failure, nerve damage, etc. When diabetes is accurately detected in its early stage , it can be effectively treated. By using Machine Learning methods, the problem can be easily detected and a solution could bearrived. Early diabetes detection and prediction can be greatly improved with machine learning (ML) approaches. When it is detected in an early stage, it can be resolved quickly. The objective of this research is to provide prediction using various supervised machine learning methods. Seven algorithms are compared with each other to figure out which is the best. The algorithms are Logistic Regression, Random Forest, Decision Tree, K-Nearest Neighbor, Support Vector Machine, Naïve Bayes, Gradient Boosting. The evaluation results stated that Logistic Regression is more accurate than other algorithms for the given data set with an accuracy of 82%. After selecting the ML model which is more accurate. A User Interface where users can enter the new data and get results was developed and the results to the user were forwarded through WhatsApp along with some suggestions and precautions.