Premiumpulse: AI-driven Insurance Cost Modeling

Abstract—Abstract—Accurate estimation of health insurance premiums is a complex task due to the nonlinear interaction of demographic, lifestyle, and medical factors. Conventional actuarial models rely on generalized risk pools and limited explanatory variables, often resulting in unfair or imprecise premium estimates. This paper proposes an AI-driven framework, termed PremiumPulse, for personalized health insurance premium prediction using supervised machine learning and ensemble regression models. The system utilizes demographic and lifestyle attributes including age, body mass index (BMI), smoking status, geographical region, gender, and number of dependents. Multiple regression algorithms—Linear Regression, Support Vector Regression (SVR), Decision Tree, Random Forest, Gradient Boosting, and XG-Boost—are evaluated using standard performance metrics such as MAE, RMSE, and R². Experiments conducted on a real-world dataset of 1,338 anonymized records using an 80–20 train-test split demonstrate that ensemble models outperform traditional regression approaches, achieving an R² score of up to 0.88 and reducing RMSE by approximately 24supports scalability, fraud-aware underwriting, and seamless integration with insurer platforms, making it suitable for real-world deployment in modern InsurTech ecosystems.