Optimizing Risk Stratification and Premium Allocation for Health Insurance Companies via Supervised Ensemble Learning

John Peter, Trinita (2025) Optimizing Risk Stratification and Premium Allocation for Health Insurance Companies via Supervised Ensemble Learning. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (3MB) | Preview
Preview	PDF (Configuration Manual) Download (730kB) | Preview

Abstract

As per the report from ‘The Irish Times‘ published on March 09, 2025, the price revision rolled out by the State’s private health insurance companies is heavily impacting both the Irish insurers and policyholders financially. Considering these growing financial pressures, the industry is now seeking innovative ways to predict and manage risks more effectively. This paper explores supervised ensemble machine learning techniques– Gradient Boosting, eXtreme Gradient Boost (XGBoost) and Random Forest that offer a promising pathway towards greater equitable risk assessment and pricing fairness. The proposed framework performs high level feature engineering to flag claims based on upper quartile threshold, bin significant categories and capture claim interaction effects by combining relevant attributes. The results demonstrate an exceptional predictive performance by XGBoost, across all instances with just 1 misclassification, strongly identifying high risk and low risk profiles, and a Reciever Operating Characteristic- Area Under Curve score (ROC) more than 95% confirming the model is well balanced and trustworthy. This engineered model serves as a prototype to streamline solutions on risk stratification, ultimately helping to stabilize insurance premiums for consumers.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Tomer, Vikas UNSPECIFIED
Uncontrolled Keywords:	Patient attributes; health variables; hyper-parameter tuning; risk anticipation; transparent policy pricing
Subjects:	Q Science > QH Natural history > QH301 Biology > Methods of research. Technique. Experimental biology > Data processing. Bioinformatics > Artificial intelligence Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Artificial intelligence R Medicine > Healthcare Industry H Social Sciences > HG Finance > Insurance
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Ciara O'Brien
Date Deposited:	25 Aug 2026 15:01
Last Modified:	25 Aug 2026 15:01
URI:	https://norma.ncirl.ie/id/eprint/9634

Actions (login required)

View Item