ARCHIVES

: Alzheimer’s disease (AD) is a progressive neurodegenerative disorder and the leading cause of dementia among the elderly. Early and accurate prediction is crucial to mitigate its social, emotional, and economic impact. This study synthesizes state-of-the-art research on the use of machine learning (ML) and deep learning (DL) techniques for the early- stage prediction of AD, drawing from clinical records, neuroimaging, and electronic health records (EHR). Feature engineering, including neuropsychological tests, MRI-derived brain volumetrics, and cerebrospinal fluid (CSF) biomarkers, plays a pivotal role in predictive accuracy. Ensemble classifiers such as Random Forest, Support Vector Machines, and XG Boost have shown robust performance, with recent models achieving up to 93% accuracy on benchmark datasets like ADNI and OASIS. Additionally, real-world EHR- based models have demonstrated the feasibility of scalable prediction in clinical settings. This work consolidates current methodologies, highlights key challenges—such as data heterogeneity and model interpretability—and proposes future directions for integrating multi-modal data and explainable AI in AD prediction.