Paper Contents
Abstract
Accurate estimation of medical insurance costs is a critical challenge for both insurance providers and policyholders due to the complex interaction of demographic, lifestyle, and health-related factors. Traditional actuarial approaches often struggle to capture non-linear relationships present in real-world healthcare data, leading to suboptimal premium pricing. This study proposes a Machine Learning-Based Medical Insurance Cost Forecasting Framework that leverages supervised learning techniques to predict individual medical insurance charges with improved accuracy and reliability.The proposed framework utilizes a structured healthcare insurance dataset containing attributes such as age, gender, body mass index (BMI), smoking status, number of dependents, and geographic region. Comprehensive data preprocessing, including encoding of categorical variables and exploratory data analysis, is performed to enhance model performance. Multiple regression-based machine learning modelsincluding Linear Regression, Ridge Regression, and Support Vector Regressionare implemented and evaluated using standard metrics such as Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and R score.Experimental results demonstrate that advanced regression models outperform traditional linear approaches, with Ridge Regression and Support Vector Regression achieving prediction accuracies of up to 82.59%, compared to 74.45% for Linear Regression. The findings confirm that lifestyle factors such as smoking status, higher BMI, and increasing age significantly influence insurance costs. The proposed framework provides a scalable and interpretable solution for accurate insurance cost forecasting, supporting fair premium calculation, risk assessment, and informed decision-making in the healthcare insurance domain.
Copyright
Copyright © 2025 Vijayalaxmi Ballolli, Mr. M.Selvam. This is an open access article distributed under the Creative Commons Attribution License.
