Peter David Hoff

Peter David Hoff is Professor of Statistical Science at Duke University, specializing in Bayesian methods for relational data, covariance estimation, and spatial modeling. Current research develops statistical frameworks for network analysis, matrix-variate data, and environmental applications.

Key methodological contributions:

• Bayesian inference for structured high-dimensional data
• Covariance estimation for matrix-variate observations
• Spatial prediction on constrained domains

Applied work includes fluorescence spectroscopy analysis for water quality monitoring and species abundance prediction. Recent publications advance source apportionment techniques and spatial covariance modeling for irregular domains.

Teaching encompasses Bayesian inference, linear models, and multivariate analysis courses at graduate levels. Professional service includes former directorship of Duke's Statistical Science graduate program.