International Research Journal of Education and Technology

Alzheimer’s disease, a progressive neurodegenerative condition with far-reaching societal consequences, highlights the urgent need for accurate predictive models to enable early diagnosis. This study focuses on evaluating the effectiveness of various machine learning techniques in forecasting Alzheimer’s, drawing insights from a comprehensive review of literature indexed in the Scopus database. By analyzing patterns across multiple studies, we identified seven commonly utilized machine learning algorithms in Alzheimer’s prediction. Among these, Support Vector Machines (SVM) demonstrated the highest predictive performance, consistently outperforming other models in terms of accuracy, sensitivity, and specificity. Random Forest also proved to be a strong contender, offering reliable results across various evaluation metrics. Our findings emphasize the crucial role of machine learning in enhancing diagnostic precision, with SVM and Random Forest standing out as particularly effective tools. This research contributes valuable guidance for the selection of suitable algorithms, supporting future efforts aimed at improving early detection and clinical intervention strategies for Alzheimer’s disease.

Keywords — Alzheimer’s diagnosis; Machine learning techniques; Predictive algorithms; Support Vector Machine (SVM); Random Forest classifier; Early detection of Alzheimer’s.