Abstract
For a variety of diagnoses, machine learning has been widely applied in healthcare. Early detection of a patient's disease can aid in better management of problems. Predictive analysis of any disease has been shown to improve the lives of those who are afflicted. Diabetes is a serious ailment that requires attention because it has negative consequences if it is ignored. Diabetes must be detected early to reverse its impact on a person's health. We used machine learning techniques on a diabetic dataset and tested the accuracy of each model in this research. For the predictive analysis, prominent factors such as glucose, pregnancies, insulin, skin thickness, age, and body mass index (bmi) are used. In this paper, the performance of various methods such as Naïve Bayes (Nbayes), k-Nearest Neighbors (kNN), Decision Trees (Dtrees), Linear Regression (LR), Vector Support Machine (svm), Linear Discriminant Analysis (LDA), Multi-Layer Perceptron (MLP) is employed in an interactive Python and the results are compared to the stacking model using Accuracy and F1 score metrics.