A Heart Disease Prediction Model using Merged XGBoost-SVM Classifier and Particle Swarm Optimization

This study aims to develop an advanced heart disease prediction model using machine learning techniques to address the global health challenge of early identification and accurate diagnosis, especially considering the projected deficits in healthcare by 2030. The proposed model integrates two powerful algorithms, XGBoost and Support Vector Machine Classifier, leveraging their complementary strengths to achieve effective classification. To enhance the model's performance, feature selection has been conducted using the LASSO method, which identifies the most relevant features, while outlier removal based on the Z score ensures data quality. StandardScaler has been utilized to standardize the data, mitigating any biases caused by differing scales. Furthermore, Particle Swarm Optimization has been employed to optimize the model's performance by fine-tuning its parameters. The reported accuracy of 91.8% underscores the model's potential to accurately predict heart disease, offering valuable support in prevention efforts. These findings highlight the significance of technological advancements in healthcare, particularly in addressing the growing demand for accurate and timely diagnoses.