Sleep Apnea and Insomnia are the sleep disorders that have a substantial influence on people’s health and quality of life. This paper provides a novel hybrid technique that combines Support Vector Machines (SVMs) and Particle Swarm Optimization (PSO) for detection of Sleep Apnea. The study makes use of the Sleep Health and Lifestyle dataset, which contains detailed information on sleep patterns such as average sleep duration, sleep quality, lifestyle factors such as daily work hours, stress level, daily exercise, steps walked, age, gender, and health metrics such as BMI and blood pressure. Feature selection from the dataset was done to improve the models’ prediction ability. Among evaluated combinations, the PSO-SVM hybrid classifier achieved the greatest accuracy of 94%, outperforming other hybrid models such as PSO-Decision Tree and PSO-Logistic Regression. Our proposed method significantly improved diagnostic accuracy, with an overall accuracy of 94%. The PSO-SVM model’s higher performance highlights the potential of hybrid techniques for sleep disorder predictions. This work emphasizes the need of combining optimization techniques with machine learning in order to create robust predictive models for healthcare applications. Our research findings indicate this approach has the potential to be a reliable alternative to traditional polysomnography, providing a more accessible and cost-effective solution for early diagnosis and management of Sleep Apnea.