Intellectual Gestational Diabetes Diagnosis System Using MLP-Whale Optimization Algorithm including Statistical Analysis

Abstract

Diabetes is one of the most widely recognized medical ailments as a silent killer in the medical services space everywhere or worldwide. It is a metabolic and persistent disease, and that indication is raising blood glucose. It leads long run to hardly harm the heart, veins, nerves, eyes, and kidneys. The causes of diabetes are hereditary, liquor utilization, smoking, obesity, activities in day to day, food habits, blood pressure, etc. Depending on the type and severity of diabetes impacts the other organs in the patient's body likewise, kidneys, heart, eyes, etc. are more prone to diseases. In this, predict diabetes using the MLP-WOA model, which is a fine-tuned weight of MLP with (WOA) Whale Optimization Algorithm. We have used a diabetes benchmark dataset taken from the UCI ML repository. We have scrutinized our model for accuracy, precision, and recall. The results have to compare against other machine learning (ML) like SVM, KNN (K-nearest neighbors), Whale Optimization Algorithm MLP, and (DTs) decision trees. We found that our MLP_WOA model performed well with an accuracy of approximately 76% than other experimental models. Also, we have tested our MLP model with other existing optimizers and observed that the WOA optimizer is giving better results.