Scalable Importance Tempering and Bayesian Variable Selection

When and Where

Tuesday, January 22, 2019 4:00 pm to 5:00 pm
Room SS1080
Sidney Smith Hall
100 St. George Street


Giacomo Zanella, Bocconi University


We propose a Monte Carlo algorithm to sample from high-dimensional probability distributions that combines Markov chain Monte Carlo and importance sampling. We provide a careful theoretical analysis, including guarantees on robustness to high-dimensionality, explicit comparison with standard schemes and computational complexity statements. Simple and concrete intuition is provided for when the novel scheme is expected to outperform standard schemes. When applied to Bayesian variable selection problems, the novel algorithm is orders of magnitude more efficient than available alternative sampling schemes and allows to perform fast and reliable fully Bayesian inferences with tens of thousand regressors. 

(Joint work with Gareth Roberts, Warwick.)



100 St. George Street