Alzheimer’s disease (AD) is a neurodegenerative disorder accountable for 60–80% of dementia cases. Early-stage recognition of the disorder is crucial for effective treatment and improved patient outcomes. Traditional methods of diagnosis, such as cognitive testing and neuroimaging, heavily depend on the physician’s judgment, leading to inconsistencies. This paper explores the use of machine learning (ML) models to automate AD diagnosis, comparing the performance of four algorithms: random forest (RF), K-nearest neighbors (KNN), naive Bayes (NB), and particle swarm optimization-tuned support vector machine (PSO-SVM). Models were tested on an open-source dataset with clinical and demographic features. Key metrics such as accuracy, precision, recall, and F1-score were used to evaluate performance. Results indicate that PSO-optimized SVM and random forest outperform other models, achieving classification accuracy exceeding 85%. PSO-tuned SVM demonstrates promising results, providing reliable and efficient decision support in medical applications.