Enhanced Prediction of Type-2 Diabetes Mellitus Using an Integrated Ensemble Framework (IEF)

Type 2 Diabetes Mellitus (T2DM) is a pressing global health challenge that significantly reduces life expectancy and diminishes quality of life. Early and accurate diagnosis is critical for avoiding the onset of serious complications. In this study, we proposed an Integrated Ensemble Framework (IEF) that utilizes a Stacking Classifier and a Fusion Model (FM) approach to enhance the accuracy of (T2DM) prediction. Initially, the outputs of two base models, the Support Vector Machine (SVM) and Gradient Boosting Classifier (GBC), are combined using the Stacking Classifier. The combined predictions are then processed through the final decision-making stage, where the Integrated Ensemble Framework is acted based on the Random Forest (RF) algorithm. The results demonstrate that IEF achieves exceptional performance, with an accuracy of 99.09%, precision of 100%, sensitivity of 98.53%, F1 score of 99.26%, and Matthews Correlation Coefficient (MCC) of 98.23%. These findings indicate that the IEF, which integrates outputs from base models, improves overall diagnostic accuracy. This IEF approach leverages the strengths of multiple algorithms, yielding more reliable predictions for T2DM diagnosis. The significant improvement in accuracy validates the potential of hybrid machine learning models in supporting early-stage diagnosis and personalized treatment planning, offering a promising solution for enhancing healthcare outcomes in T2DM patients.