Vol. Quarter 1, 2025

E-Forum

We demonstrate a synthetic data generation algorithm that balances privacy and utility, preserving statistical relationships while enabling regulatory compliance for secure, accurate insurance modeling.

This paper explores credibility theory using data augmentation, providing an intuitive interpretation of the “K” parameter in Bayesian approximation and alternative applications in ratemaking with GLMs and severity curve-fitting.

This paper is an introduction and guide to agent-based modeling (ABM), and it includes several case studies to illustrate their utility.

ReSurv is an R package for estimating feature-dependent development factors using individual claims data, enhancing non-life reserving beyond traditional chain-ladder methods. This paper provides a practitioner-focused guide to its application.