Early Identification of Alzheimer’s Disease Through Cognitive Facets

This paper presents an efficient and non-invasive machine learning-based framework for the early detection of Alzheimer's disease using cognitive features.Early detection of Alzheimer's disease (AD) is crucial for enabling timely clinical intervention and improving patient outcomes.The proposed system employs a hybrid feature selection technique combining Neighbourhood Component Analysis and Correlation-based Filtration to identify the most discriminative cognitive attributes.The optimized features are then classified using a robust votingbased ensemble learning model that integrates KNN, Random Forest, AdaBoost, XGBoost, Naï ve Bayes, Logistic Regression, and Decision Tree classifiers.A majority voting strategy is used to derive the final prediction, enhancing stability and generalization.Experimental results demonstrate that the ensemble model achieves an accuracy exceeding 93%, significantly outperforming individual non-ensemble classifiers, which achieved a maximum accuracy of 83%.The system is implemented through a user-friendly web interface supporting dataset upload, feature selection, model training, and real-time disease prediction.By leveraging cognitive assessments instead of costly neuroimaging techniques, the proposed framework offers an accessible, scalable, and clinically interpretable solution for early-stage Alzheimer's diagnosis.