Generative Models and Uncertainty Quantification

Copenhagen, October 9 and 10, 2019

Generative models and uncertainty quantification lie at the heart of Bayesian modelling and inference. At this small meeting, we discuss recent developments within the field. The meeting is deliberately kept small in order to ensure that discussion remains honest, lively and interesting. Attendance is, thus, mostly by invitation, but one can apply to join (see below).

Confirmed Speakers

Max Welling
Professor in Machine Learning at the University of Amsterdam and a VP Technologies at Qualcomm.

Aki Vehtari
Associate professor in computational probabilistic modeling at Aalto University; visiting professor at Technical University of Denmark (DTU).

Will Grathwohl
PhD student at the University of Toronto.

Mark van der Wilk
Machine Learning Researcher PROWLER.io.

José Miguel Hernández Lobato
University Lecturer in Machine Learning, University of Cambridge.

Rianne van den Berg
Research Scientist at Google AI, Amsterdam.

Sebastian Nowozin
Machine learning researcher in the Brain team at Google AI, Berlin.

Daniel Hernández-Lobato
Lecturer of Computer Science at Universidad Autónoma de Madrid.

Andriy Mnih
Research scientist at DeepMind.

Tom Rainforth
Researcher in statistical machine learning at the University of Oxford.

Casper Kaae Sønderby
Research Scientist in Google Brains Amsterdam Lab.

Julie Josse
Professor of Statistics at Ecole Polytechnique (CMAP) and XPOP INRIA team.

Silvia Chiappa
Staff Research Scientist in Machine Learning at DeepMind.

Program

Wednesday (Oct 9)		Thursday (Oct 10)
Session 1 Advances in deep generative models Chair: Søren Hauberg		Session 3 Bayesian neural networks Chair: Wouter Boomsma
8:55-9:00	Opening remarks
9:00-10:00	Max Welling The Functional Process VAE	09:00-09:40	Sebastian Nowozin Practical Bayesian Neural Networks: Three Obstacles
10:00-10:40	Andriy Mnih Resampled Priors for Variational Autoencoders	09:40-10:40	Casper Kaae Sønderby Bayesian Inference for Large Scale Image Classification
10:40-11:00	Coffee	10:40-11:00	Coffee
11:00-11:40	Rianne van den Berg Normalizing flows for discrete data	Session 4 Generative models for downstream tasks Chair: Ole Winther
11:40-12:40	Tom Rainforth Disentangling Disentanglement in Variational Autoencoders	11:00-12:00	José Miguel Hernández-Lobato Advances in Compression via Probabilistic Machine Learning
Session 2 Bayesian and causal inference Chair: Pierre-Alexandre Mattei		12:00-13:00	Lunch
12:40-13:40	Lunch	13:00-14:00	Mark van der Wilk Sampling for data augmentation: generative models vs invariances
13:40-14:40	Aki Vehtari Bayesian workflow	14:00-15:00	Will Grathwohl The many virtues of Incorporating energy-based generative models into discriminative learning
14:40-15:00	Coffee	14:40-15:00	Coffee
15:00-15:40	Daniel Hernández-Lobato Adversarial Alpha Divergence Minimization for Bayesian Approximate Inference
15:40-16:20	Silvia Chiappa Causal inference and fairness
16:20-18:00	Poster Session
18:30-22:00	Conference Dinner at Toldboden (directions)

Registration

If you would like to attend this meeting, but have yet to recieve an invitation, then fill out the following Google form. Note that seating is limited, so we cannot guarantee a ticket.

Venue

The workshop takes place at:

Pier47
Langelinie Allé 47
2100 Copenhagen Ø, Denmark

Pier47

Organizers

The conference is jointly organized by:

We are grateful for funding from the Villum Foundation (grant 15334) and the European Research Council (ERC) under the European Union's Horizon 2020 research and innovation programme (grant agreement number 757360).