Bayesian Monte Carlo
Ghahramani, Zoubin, Rasmussen, Carl E.
–Neural Information Processing Systems
We investigate Bayesian alternatives to classical Monte Carlo methods for evaluating integrals. Bayesian Monte Carlo (BMC) allows the incorporation ofprior knowledge, such as smoothness of the integrand, into the estimation. In a simple problem we show that this outperforms any classical importance sampling method. We also attempt more challenging multidimensionalintegrals involved in computing marginal likelihoods ofstatistical models (a.k.a.
Neural Information Processing Systems
Dec-31-2003