Anticipating Diabetes using Fusion-Ensemble Machine Learning Techniques

Abstract

Diagnosis of diabetes is still a complex task for most of the already existing machine learning methods. The objective of the study is to develop a decision support system for predicting the probability of diabetes disease. By using machine learning algorithms and considering fuzzy logic to handle uncertainty, a decision support system can be developed for predicting the probability of having diabetes in the given dataset. Diabetes patients will increase day by day; In current research, several algorithms have been used to predict diabetes. These are major issues for diabetes. Using these technologies, A model will be implemented to predict diabetes using different algorithms with proper comparison and find the best algorithm to predict diabetes. The prediction model is made using 11 classification algorithms from Ski learn, and their accuracies are compared. The expected result is that one of the best algorithms can be obtained for making a diabetes detection model. The fusion (ensemble) model is used for diabetes prediction, intended to improve the accuracy of classification. and use two algorithms to apply the fusion (ensemble) model, which picks the maximum accuracy of the list of classifiers with the rest of the classifiers. The source code is publicly accessible at https://github.com/diabetes-mellitus-implementation.