ENHANCING EARLY DIAGNOSIS OF HEART DISEASES USING MACHINE LEARNING MODELS AND PREDICTIVE ANALYTICS

Abstract

Cardiovascular diseases (CVDs) remain one of the leading causes of mortality worldwide, underscoring the urgent need for accurate and proactive prediction models. This study presents a novel machine learning-based approach for heart disease prediction, emphasizing the integrated analysis of clinical data and lifestyle factors that are often underrepresented in traditional models. A hybrid framework was developed, combining advanced feature selection techniques with ensemble learning algorithms to enhance both predictive accuracy and model robustness. A key contribution of this work lies in a comprehensive feature engineering strategy that captures interaction effects among lifestyle variables—such as diet, physical activity, and stress levels—alongside conventional clinical indicators. The model was trained and evaluated on a diverse patient dataset to ensure broader generalizability across populations. Experimental results demonstrate a notable improvement over existing approaches, achieving an Area Under the Curve (AUC) of 0.90, with strong capability in identifying high-risk individuals. In addition, an in-depth feature importance analysis was conducted to identify the most influential risk factors, providing meaningful insights to support clinical decision-making and personalized healthcare interventions. The findings highlight the critical role of lifestyle factors in cardiovascular risk assessment and demonstrate the effectiveness of combining them with clinical data. Overall, this research contributes to the advancement of heart disease prediction by offering a more comprehensive, accurate, and practical framework for early diagnosis and preventive care.