Overview
Kosuke Imai is a professor in the Department of Politics at Princeton University. He works at the intersection of political methodology and Bayesian computation. He is co-developer (with David van Dyk) of the marginal data augmentation algorithm for the multinomial probit model that dominates all prior Bayesian methods on the criteria of prior interpretability, computational speed, and simplicity.
Key Contributions
- Imai and van Dyk (2005): New MCMC algorithms for the multinomial probit model via marginal data augmentation; places prior directly on identifiable (β, Σ); Scheme 1 marginalization improves convergence over all prior methods; corrects Nobile (1998) acceptance probability error; C-code and R interface publicly available. See Multinomial Probit and Imai-van Dyk (2005).
Related