<![CDATA[Alzheimer's disease is the main reason people develop dementia around the world. However, diagnostic methods that use medical imaging, like MRI scans, are usually costly and not easily available, especially in hospitals and basic healthcare centers. This study is focused on creating and evaluating different classification models using clinical data, and it compares how well the Stacking Ensemble method and the Naïve Bayes method perform. The Stacking Ensemble model uses Random Forest, XGBoost, and Support Vector Machine as the base models, and Logistic Regression acts as the meta-learner.Gaussian Naïve Bayes is also used as the baseline model. The dataset came from Kaggle and includes 2,149 patient records with 32 predictor features after some cleaning and preparation. The data were split into 80% for training and 20% for testing using a stratified method. The model's performance was checked on the testing data and confirmed using 5-fold stratified cross-validation. The experimental results show that the Stacking Ensemble method got an accuracy of 94.65%, precision of 93.88%, recall of 90.79%, and an F1-score of 92.31%. These results do much better than the Naïve Bayes method, which only reached an accuracy of 77.21%. Moreover, cross-validation showed that the proposed model is reliable, achieving an average accuracy of 95.11% for the Stacking Ensemble. These results show that the Stacking Ensemble approach works better and more reliably than the Naïve Bayes model when classifying Alzheimer's disease based on clinical data.]]>
Copyrights © 2026
