Overview
Alan E. Gelfand is a Bayesian statistician, formerly Professor at the Department of Statistics, University of Connecticut, and later at Duke University. He is best known for co-introducing the Gibbs sampler to mainstream Bayesian statistics in the landmark Gelfand-Smith (1990) paper, which triggered the MCMC revolution in Bayesian computation; he is also a leading figure in Bayesian spatial statistics.
Key Contributions / Features
- Gelfand-Smith (1990) — "Sampling-Based Approaches to Calculating Marginal Densities," Journal of the American Statistical Association 85(410): 398–409. Co-introduced Gibbs sampling from image processing to standard statistical inference; proved equivalence with the Tanner-Wong substitution algorithm; established the Rao-Blackwell density estimator; demonstrated applicability across hierarchical Bayesian models with conjugate full conditionals.
- Neelon-Gelfand (2014) — multivariate spatial mixture model for areal data: CAR-smoothed finite mixtures for correlated continuous outcomes. See Conditional Autoregressive Model.
- Gelfand-Vounatsou (2003) — proper multivariate CAR models with spatial autoregression parameters for hierarchical spatial data. See Gelfand-Vounatsou (2003).
Related