Modeling the Determinants of Satisfaction with Health Care Services in Türkiye Using Machine Learning

Purpose: Patient satisfaction is a fundamental indicator for evaluating the quality of healthcare services and is of great importance for the sustainability of healthcare systems.Methods: In this study, machine learning methods were used to identify factors affecting individuals' levels of satisfaction with healthcare services and to predict their satisfaction levels. The Turkish Statistical Institute (TÜİK) life satisfaction dataset was used; the data were split into training (70%) and test (30%) sets, and the SMOTE method was applied to address class imbalance. Results: Logistic Regression, SVM, Decision Trees, Random Forest, and Naive Bayes models were tested, and Logistic Regression was determined to be the most suitable model. According to the findings, married individuals with a high level of education and those who are employed are more satisfied. In contrast, hygiene issues, insufficient healthcare personnel, long waiting times, and high insurance costs reduce satisfaction. Conclusion: Ultimately, machine learning approaches are practical for analyzing healthcare satisfaction.