Simulation-based stacking

arXiv:2310.17009v2 Announce Type: replace-cross
Abstract: Simulation-based inference has been popular for amortized Bayesian computation. It is typical to have more than one posterior approximation, from different inference algorithms, different architectures, or simply the randomness of initialization and stochastic gradients. With a consistency guarantee, we present a general posterior stacking framework to make use of all available approximations. Our stacking method is able to combine densities, simulation draws, confidence intervals, and moments, and address the overall precision, calibration, coverage, and bias …

abstract algorithms approximation architectures arxiv bayesian computation cs.lg framework general inference popular posterior randomness simulation stat.co stat.me stochastic type