Dear all, the first One World Approximate Bayesian Inference (OWABI) seminar www.warwick.ac.uk/owabi<http://www.warwick.ac.uk/owabi> of the year is quickly approaching! Our next speaker is Louis Sharrock<https://louissharrock.github.io/> <https://louissharrock.github.io/> (University College London) who will talk about "Sequential Neural Score Estimation: Likelihood-free inference with conditional score base diffusion models" on Thursday the 29th January at 11am UK time, with an abstract reported below. The talk will be streamed on the OWABI MS Team channel General | OWABI Seminar: One World Approximate Bayesian Inference Seminar | Microsoft Teams<https://teams.microsoft.com/l/team/19%3AdhZ_4e_XLNJzCXPAMzTvT6BZ5KShEETkd_wtTY52VI81%40thread.tacv2/conversations?groupId=9c061d11-f88c-4cee-938f-bf40e7393879&tenantId=09bacfbd-47ef-4465-9265-3546f2eaf6bc>, at the link below Join: https://teams.microsoft.com/meet/38899039621449?p=nCtiqu8zYnZNL2LGR8 Meeting ID: 388 990 396 214 49 Passcode: f3Dh22XL We are looking forward to seeing you at the next OWABI seminar, Best, Massimiliano on the behalf of the OWABI seminar organisers Speaker: Louis Sharrock (University College London) Title: Sequential Neural Score Estimation: Likelihood-free inference with conditional score base diffusion models Abstract: We introduce Sequential Neural Posterior Score Estimation (SNPSE), a score-based method for Bayesian inference in simulator-based models. Our method, inspired by the remarkable success of score-based methods in generative modelling, leverages conditional score-based diffusion models to generate samples from the posterior distribution of interest. The model is trained using an objective function which directly estimates the score of the posterior. We embed the model into a sequential training procedure, which guides simulations using the current approximation of the posterior at the observation of interest, thereby reducing the simulation cost. We also introduce several alternative sequential approaches, and discuss their relative merits. We then validate our method, as well as its amortised, non-sequential, variant on several numerical examples, demonstrating comparable or superior performance to existing state-of-the-art methods such as Sequential Neural Posterior Estimation (SNPE). Keywords: diffusion models, simulation based inference, sequential methods. Reference: L. Sharrock, J. Simons, S. Liu, M. Beaumont, Sequential Neural Score Estimation: Likelihood-Free Inference with Conditional Score Based Diffusion Models<https://proceedings.mlr.press/v235/sharrock24a.html>. PLMR, 235, 44565-44602, 2024. ------ Dr. Massimiliano Tamborrino Reader (Associate Professor) and WIHEA Fellow Department of Statistics University of Warwick https://warwick.ac.uk/tamborrino