Program of the workshop

Schedule

The first day of the event consists in four introductory courses, tackling the basics of their respective topics. The other days of the event are designed to allow invited speakers to present and discuss ongoing research and state of the art results.
Most talks are recorded and videos are available on LIKE22’s Switchtube channel. You can also access the recorded talks below.

Monday 10th January 2022

Time	Speaker	Title and content
09:00-10:30	Athénaïs Gautier & David Ginsbourger	Flexible, probabilistic function modelling with Gaussian Processes Click here to expand and see the recording. Download the slides Download the notebook (please, remember to save the notebook with the ".ipynb" extension)
11:00-12:30	Dario Azzimonti & Cédric Travelletti	Sequential design of experiments with Gaussian Process models Click here to expand and see the recording. Download the slides (Part 1) Download the slides (Part 2)
13:30-15:00	Soham Sarkar	Kernel methods: past, present, future Click here to expand and see the recording. Download the slides
15:30-17:00	ST John	Gaussian processes for non-Gaussian likelihoods Click here to expand and see the recording. Download the slides

Tuesday 11th January 2022

Time	Speaker	Title and content
09:00-10:30	Krikamol Muandet	Kernel Mean Embedding with Applications in Deconfounded Causal Learning Click here to expand and see the recording. Download the slides
11:00-11:45	Tomasz Kacprzak	Adventures in inference in high dimensional spaces with heavy simulations in the field of cosmology. ◷ Recording not available yet, please come back later.
11:45-12:30	Niklas Wahlström	Linearly and nonlinearly constrained Gaussian processes Click here to expand and see the recording.
15:00-16:00		Video presentations from early-stage researchers. Watch their videos here or directly on Switchtube
16:00-17:30	Peter Frazier	Grey-Box Bayesian Optimization Click here to expand and see the recording. Download the slides

Wednesday 12th January 2022

Time	Speaker	Title and content
09:00-10:30	Michael Gutmann	Accelerating Approximate Bayesian Computation with Kernels and Decision Making under Uncertainty Click here to expand and see the recording.
11:00-11:45	Richard Wilkinson	Adjoint-aided inference of Gaussian process driven differential equations Click here to expand and see the recording.
11:45-12:30	Chris Oates	Robust Generalised Bayesian Inference for Intractable Likelihoods Click here to expand and see the recording.
14:00-14:45	Zoltan Szabo	Continuous Emotion Transfer using RKHSs ☒ This talk was not recorded. Download the slides
14:45-15:30	Amandine Marrel	New advances in sensitivity analysis based on HSIC dependence measures ☒ This talk was not recorded. Download the slides
16:00-16:45	Danica Sutherland	Better deep learning (sometimes) by learning kernel mean embeddings Click here to expand and see the recording.

Thursday 13th January 2022

Time	Speaker	Title and content
09:00-10:15	Florence d'Alché-Buc	Learning to predict complex outputs: a kernel view Click here to expand and see the recording.
10:45-11:30	George Wynne	A Spectral View of Kernel Stein Discrepancy: Unlocking Infinite Dimensions Click here to expand and see the recording.
11:30-12:15	Johanna Ziegel	Kernel scores: A versatile class of proper scoring rules for evaluating probabilistic forecasts Click here to expand and see the recording. Download the slides
13:30-14:45	José Miguel Hernández-Lobato	Molecule optimization with deep generative models Click here to expand and see the recording.
14:45-15:30		Q&A with selected early-stage researchers. ☒ This recording is not available.
16:00-17:15	Andrew Gordon Wilson	How should we build scalable Gaussian processes? Click here to expand and see the recording.

Friday 14th January 2022

Time	Speaker	Title and content
09:00-10:15	Mark van der Wilk	Approximations, Inductive Biases, and their Connections in Gaussian Processes Click here to expand and see the recording.
10:45-11:30	Dario Azzimonti	Skew Gaussian Processes for classification, preference and mixed problems Click here to expand and see the recording.
11:30-12:15	Carl Henrik Ek	Modulated Surrogates for Bayesian Optimisation Click here to expand and see the recording.
12:15-12:30		Closing of LIKE22.