Dear all,

You are cordially invited to an academic seminar to be delivered by Prof. Jeremias Knoblauch on October 26 (Thursday). Please find the details as follows.

Seminar Information

Topic: Optimisation-centric Generalisations of Bayesian Inference

Time & Date: 21:00-22:00, Beijing Time, October 26 (Thursday)

Online Meeting:

Zoom Meeting Link: https://cuhk-edu-cn.zoom.us/j/91982083649?pwd=a0RNbkpQYllYYy9Lb0VCQklYamYrdz09

Meeting ID: 919 8208 3649

Passcode: 123456

Speaker: Prof. Jeremias Knoblauch, University College London

Host:  Prof. Feng YIN, The Chinese University of Hong Kong, Shenzhen

Abstract: I summarize a recent line of research and advocate for an optimization-centric generalisation of Bayesian inference. The main thrust of this argument relieson identifying the tension between the assumptions motivating the Bayesian posterior and the realities of modern Bayesian Machine Learning. Our generalisation is a useful conceptual device, but also has methodological merit: it can address various challenges that arise when the standard Bayesian paradigm is deployed in a Machine Learning context—including robustness to model misspecification, robustness to poorly chosen priors, and inference in intractable models.

Biography

Jeremias Knoblauch is an Assistant Professor and Engineering & Physical Sciences Research Council (EPSRC) Fellow at the department of statistical science at University College London (UCL). Prior to that, he was a Biometrika Fellow. Before his time at UCL, Jeremias obtained his Ph.D. as part of the Oxford-Warwick Statistics Programme, as part of which he became the first UK-based Facebook Fellow.