Overview
David P. M. Scollnik is an actuary and statistician at the University of Calgary (Department of Mathematics and Statistics). Known primarily within the actuarial community for introducing Gibbs sampling and MCMC methods to casualty actuarial practice via a tutorial paper in the CAS Proceedings (~1996). Also contributed Bayesian simultaneous equations models for insurance ratemaking.
Key Contributions
- Scollnik (~1996): 52-page CAS Proceedings tutorial on MCMC and Gibbs sampling. Four worked examples: Generalized Pareto mixture, Beta-Poisson-Binomial portfolio model, data augmentation for grouped Pareto data, and hierarchical credibility for workers compensation (HNM, HFLPM, VSHNM models on 133 occupation classes). Introduced Gelman-Rubin and Raftery-Lewis convergence diagnostics to the actuarial audience; promoted BUGS software.
- Scollnik (1995/1996): Bayesian simultaneous equations model for insurance ratemaking via Gibbs sampler (cited in the tutorial).
Related