Identifying Vulnerable Populations for Health Insurance Claims Using Cluster Analysis

This presentation explores how K-means clustering can categorize health-insured populations based on demographic, clinical, and behavioral traits to pinpoint vulnerable groups likely to have high healthcare expenses. By using the Elbow Method and Silhouette Score to validate clusters, distinct profiles were identified, including high-risk populations with issues such as obesity, hypertension, and smoking rates. The results underscore the importance of unsupervised learning in enhancing risk stratification, assisting insurers with premium adjustments, directing preventive measures for providers, and informing fair healthcare policy decisions

To explain how K-means clustering can be applied to health insurance data to identify high-risk populations

To distinguish between traditional actuarial approaches and unsupervised learning methods for healthcare risk stratification

To utilize clustering insights for developing targeted strategies that enhance cost management and promote healthcare equity