Overview
Arts & Sciences Professor of Statistics at Duke University (Institute of Statistics and Decision Sciences). One of the leading figures in Bayesian forecasting, dynamic linear models, and Bayesian computation. Co-author with Jeff Harrison of Bayesian Forecasting and Dynamic Models (Springer, 2nd ed. 1997), the standard reference for the DLM framework. Research spans dynamic factor models, stochastic volatility, particle methods, and Bayesian variable selection. Earlier work on sequential Bayesian methods (West 1993) and variance discounting (West-Harrison 1997) underpins the sequential filtering approach in Aguilar-West (2000).
Key Contributions / Features
- West and Harrison (1997) — Bayesian Forecasting and Dynamic Models, 2nd ed. Springer. Definitive reference for the DLM framework; forward-backward filtering, discount variance matrices (St=(1−at)St−1+atytyt′), and sequential Bayesian updating. Backward smoothing formulae (pp. 608–609) used directly as the variance-discounting baseline in Aguilar-West (2000).
- Aguilar and West (2000) — "Bayesian Dynamic Factor Models and Portfolio Allocation," JBES 18(3): 338–357. Joint with Omar Aguilar; k-factor SV model with VAR(1) log-volatility dynamics and MCMC + APF inference; 6-currency exchange-rate portfolio application; demonstrated advantage of structured volatility modelling over variance discounting.
- Lopes and West (1998) — Working paper extending MCMC to static Bayesian factor analysis with joint inference on the number of factors k; referenced in Aguilar-West as the precursor to extending k-inference to dynamic SV factor models.
Related