Overview
Ivan Jeliazkov is an economist who was a doctoral student at Washington University in St. Louis (Olin Business School) when he co-authored the Chib-Jeliazkov (2001) paper with Siddhartha Chib. The paper extended the Chib (1995) marginal likelihood identity to samplers containing Metropolis-Hastings steps, making universal marginal likelihood estimation feasible for any MCMC-fit model.
Key Contributions / Features
- Chib-Jeliazkov (2001): Co-authored with Siddhartha Chib; derived the CJ posterior ordinate estimator from the local reversibility of M-H subkernels (eq. 9–10); extended the method to multi-block samplers with latent variables (eq. 11–18); established the link between MCMC sampler efficiency and NSE of the marginal likelihood estimate. See Marginal Data Density.
Related