Approximation of Zero-Inflated Poisson Credibility Premium via Variational Bayes Approach

While both zero-inflation and the unobserved heterogeneity in risks are prevalent issues modeling insurance claim counts, determination of Bayesian credibility premium counts with these features often demanding due to high computational costs associated a use MCMC. This article explores way approximate for claims frequency that follows zero-inflated Poisson distribution via variational Bayes approach. Unlike many existing industry benchmarks, proposed method enables companies capture policyholders simultaneously modest computation costs. A simulation study an empirical analysis using LGPIF dataset were conducted it turned out outperforms benchmarks terms prediction performances time. Such results support applicability posterior ratemaking practices.