Answers are grounded in this paper and cite their sources.
Health claim insurance companies are conditioned by two level risks, In the one hand under-provisioning and over-provisioning, In the other hand information asymmetry and adverse selection. Conventional methods fail to reflect the true risk class of the insurers, which leads to biased strategic decisions. This paper proposes a novel rik-profiling framework for health claim insurance that balances the provisioning and infers the likely profil of the individual. The proposed framework is built in two-level structure. The first level comprises a regression approach to face the risk of under/over provisioning in which we used a list of regressors outperformed by Extratrees (R 2 =0.9649). The second level contains two sub-levels, a clustering level based on K-means that defined three risk clusters and a Classification level using different statistical/machine learning models where CART was the best profils’ explainer (accuracy=0.999862).
No takes yet. Share an insight, caveat, or question.
Saoudi et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: