4 papers
Amortized Bayesian Workflow
Chengkun Li, Aki Vehtari, Paul-Christian Bürkner +3
Bayesian inference often faces a trade-off between computational speed and sampling accuracy. We propose an adaptive workflow that integrates rapid amortized inference with gold-st…
Stacking Variational Bayesian Monte Carlo
Francesco Silvestrin, Chengkun Li, Luigi Acerbi
Approximate Bayesian inference for models with computationally expensive, black-box likelihoods poses a significant challenge, especially when the posterior distribution is complex…
Fast post-process Bayesian inference with Variational Sparse Bayesian Quadrature
Chengkun Li, Grégoire Clarté, Martin Jørgensen +1
In applied Bayesian inference scenarios, users may have access to a large number of pre-existing model evaluations, for example from maximum-a-posteriori (MAP) optimization runs. H…
Normalizing Flow Regression for Bayesian Inference with Offline Likelihood Evaluations
Chengkun Li, Bobby Huggins, Petrus Mikkola +1
Bayesian inference with computationally expensive likelihood evaluations remains a significant challenge in many scientific domains. We propose normalizing flow regression (NFR), a…