Celia Christy Merlinda Tantono, Aqilla Dheanya Lucetta, Putri Felicia, Edwin Setiawan Nugraha

APPLICATION OF LOGISTIC REGRESSION FOR HEALTH INSURANCE PREMIUM CLASSIFICATION

Introduction

Application of logistic regression for health insurance premium classification . Classify health insurance premiums (high/low) using logistic regression based on medical and demographic predictors. Identifies Age and Weight as significant, achieving 97% AUC and 95% accuracy for risk assessment.

Abstract

This study analyzes individual health insurance data using logistic regression to classify premiums into high and low categories based on ten medical and demographic predictors. By transforming a continuous premium variable into binary classes, high and low premiums, this research evaluates the significant factors influencing premium pricing decisions. Logistic regression was selected for its ability to model binary outcomes and estimate the likelihood that the customer belongs to the high premium category. To evaluate the significance of the predictors and the overall model fit, the Likelihood Ratio Test and Wald test were performed, identifying Age and Weight as significant predictors affecting premium classification. The final logistic regression model has an excellent ability to predict, given the area under the curve (AUC) of 0.97 and a 95% accuracy. These results show how well logistic regression works to enhance risk classification and support data driven policy modifications in insurance underwriting procedures.

Full Text

Comments

You need to be logged in to post a comment.