Evaluating Determinants of Health Insurance Premiums Using Advanced Multiple Linear Regression Techniques

The decision to purchase health insurance policies is a common strategy to manage the escalating costs of medical treatment. This study aims to statistically identify the key factors determining health insurance premium prices. A variety of methods were applied, including Ordinary Least Square Regression (OLS), Ridge Regression, Lasso Regression, and Support Vector Regression (SVR), to determine the most suitable model for predicting premium costs. The analysis focused on multiple factors such as age, gender, Body Mass Index (BMI), number of children, smoking status, and region. OSL analysis revealed that age, BMI, number of children, and smoking status positively affect the value of health insurance. Also, it has been shown that the prices vary with respect to regions, while gender is not a significant determinant of charge. Smoking status has the highest impact, while age is the least, and BMI and region are almost the same. Among the methods tested, Support Vector Regression (SVR) demonstrated the lowest Root Mean Square Error (RMSE) of 0.84, indicating it provided the best fit for predicting health insurance costs based on these variables. The findings highlight SVR as an effective tool for estimating health insurance premiums, offering insights into how various personal and demographic factors influence the cost. The results contribute to a deeper understanding of the key drivers that help customers anticipate future costs and allow insurance companies to adopt a more precise tool for pricing more tailored, data-driven premiums.