{"success":1,"msg":"","color":"rgb(28, 35, 49)","title":"<b>Nonparametric Bayesian inference for reversible multi-dimensional  diffusions<\/b>","description":"webinar","title2":"","start":"2021-05-21 15:00","end":"2021-05-21 16:00","responsable":"<a href=\"mailto:b.t.szabo@vu.nl\">Botond Szabo <i class=\"fa fa-envelope\"><\/i><\/a>","speaker":"Matteo Giordano (Cambridge)","id":"41","type":"webinar","timezone":"Europe\/Amsterdam","activity":"Zoom meeting:\r\n\r\nlink: https:\/\/vu-live.zoom.us\/j\/95319105586?pwd=MXlIWXIvNjcwQ0k1YzErMlkyTzNZQT09\r\nMeeting ID: 953 1910 5586\r\nPasscode: 645009","abstract":"The talk will present frequentist asymptotic results for \r\nnonparametric Bayesian models of reversible multi-dimensional diffusions \r\nwith periodic drift. Assuming continuous observation paths, \r\nreversibility is exploited to prove a general posterior contraction rate \r\ntheorem for the drift gradient vector field under \r\napproximation-theoretic conditions on the induced prior for the \r\ninvariant measure. The general theorem is applied to Gaussian priors and \r\np-exponential priors, which are shown to converge to the truth at the \r\nminimax optimal rate over Sobolev smoothness classes in any dimension.\r\n\r\nJoint work with Kolyan Ray (Imperial College London)."}