DIAGNOSING DIABETES WITH MACHINE LEARNING TECHNIQUES

The rate of diabetes is rapidly increasing worldwide. Early detection can help prevent or delay the onset by initiating lifestyle changes and taking appropriate preventive measures. Until now, prediabetes type 2 have proved to be early problems. There a need for easy, rapid, accurate diagnostic tools diagnosis in this context. Machine learning algorithms diagnose diseases early. Numerous studies are being conducted improve speed, performance, reliability, accuracy diagnosing with these methods particular disease. This study aims predict whether patient has based on measurements dataset from National Institute Diabetes Digestive Kidney Diseases. Eight different variables belonging patients were selected as input variable, it was estimated had not. Of 768 records examined, 500 (65.1%) healthy, 268 (34.9%) diabetes. Ten machine been applied diabetic status. most successful method Random Forest algorithm 90.1% accuracy. Accuracy percentages other also between 89% 81%. describes highly prediction tool finding model identified may helpful diagnosis.