Experimental Evaluation of Diabetes Mellitus Prediction Scheme Based on Enhanced Machine Learning Strategy

As the leading cause of death and morbidity among non-communicable diseases, diabetes impacts millions of people worldwide. Diabetes can develop for a variety of reasons, including being overweight, having high cholesterol levels, having a family history of the disease, not getting enough exercise, eating poorly, etc. A variety of human organs are vulnerable to the increasingly common diabetes mellitus. Diabetic susceptibility is complex and depends on several factors. These include genetics, lifestyle choices, environmental factors, and family medical history. A collection of diseases affecting the body's use of glucose is together known as diabetes mellitus. The causes of diabetes vary among the various types. However, it will cause hyperglycemia in those with any kind of diabetes. Type 1 as well as type 2 diabetes will exist separately. As a society, we will benefit much from early prediction.It will offer a safe path for human living. Chronic damage and malfunction of several tissues, including the retina, livers, heart and nerves, which can be caused by long-term high blood sugar and hence, it is extremely crucial to be able to predict diabetes in its early stages. Six hundred people with diabetes or at risk for developing the disease, ranging in age from twenty-five to eighty-five, were used as training data for various machine learning algorithms, including PLSDM and SVM, in this research. The suggested method outperforms the alternatives when comparing recognition and classification accuracy.