van der Schaar Lab

Lab’s pioneering projects on show at NeurIPS 2020

The van der Schaar Lab’s research leadership and very latest projects will be highly visible at NeurIPS 2020 (widely considered the world’s largest AI and machine learning research conference). The lab will publish a record 9 papers at the conference, while Mihaela van der Schaar will deliver keynotes at two workshops. Several lab members, including Mihaela, will also give a presentation on the lab’s Hide-and-seek Privacy Challenge, which was held as part of the conference’s competition track.

Keynote at Women in Machine Learning Workshop

Mihaela will be giving a keynote at the 15th Women in Machine Learning Workshop (WiML 2020), which will be held alongside the virtual NeurIPS conference. Her talk will take place on December 9 at 11:00 GMT (full list of time zones here).

The full program and event details can be found here.

Keynote at NeurIPS Europe meetup on Bayesian Deep Learning

Mihaela will give a keynote entitled Bayesian Uncertainty Estimation under Covariate Shift: Application to Cross-population Clinical Prognosis at the NeurIPS Europe meetup on Bayesian Deep Learning on December 10 at 11:30 GMT (full list of time zones here).

You can learn more about this workshop here.

Presentation on Hide-and-seek Privacy Challenge

Several members of the van der Schaar Lab, including Mihaela, will give a presentation on the lab’s recent Hide-and-Seek Privacy Challenge, a novel two-tracked competition to explore the meaning and limitations of data privacy through the use of synthetic data.

The goal of the competition, which ran as part of the NeurIPS 2020 competition track, was to identify methods capable of generating private synthetic data. The challenge was administered by the van der Schaar Lab with support from the University of Cambridge, Microsoft Research, and Amsterdam UMC

This will take place on December 11; details can be found here.

Total of 9 papers selected for publication

The lab has set a new group record for representation at NeurIPS 2020, with a total of 9 papers accepted for publication.

This is an unprecedented achievement for the lab, and demonstrates the diverse strengths of its small research team. The papers cover diverse topics, such as interpretability, uncertainty quantification, causal inference, and imitation learning. Applications in healthcare are similarly broad, ranging from treatment effect estimation to predicting the impact of COVID-19 spread prevention policies.

Titles, authors and abstracts for all 9 selected papers are given below.

When and How to Lift the Lockdown? Global COVID-19 Scenario Analysis and Policy Assessment using Compartmental Gaussian Processes

Zhaozhi Qian, Ahmed Alaa, Mihaela van der Schaar

Abstract

Robust Recursive Partitioning for Heterogeneous Treatment Effects with Uncertainty Quantification

Hyun-Suk Lee, Yao Zhang, William Zame, Cong Shen, Jang-Won Lee, Mihaela van der Schaar

Abstract

VIME: Extending the Success of Self- and Semi-supervised Learning to Tabular Domain

Jinsung Yoon, Yao Zhang, James Jordon, Mihaela van der Schaar

Abstract

OrganITE: Optimal transplant donor organ offering using an individual treatment effect

Jeroen Berrevoets, James Jordon, Ioana Bica, Alexander Gimson, Mihaela van der Schaar

Abstract

CASTLE: Regularization via Auxiliary Causal Graph Discovery

Trent Kyono, Yao Zhang, Mihaela van der Schaar

Abstract

Learning outside the Black-Box: The pursuit of interpretable models

Jonathan Crabbe, Yao Zhang, William Zame, Mihaela van der Schaar

Abstract

Estimating the Effects of Continuous-valued Interventions using Generative Adversarial Networks

Ioana Bica, James Jordon, Mihaela van der Schaar

Abstract

Gradient Regularized V-Learning for Dynamic Treatment Regimes

Yao Zhang, Mihaela van der Schaar

Abstract

Strictly Batch Imitation Learning by Energy-based Distribution Matching

Daniel Jarrett, Ioana Bica, Mihaela van der Schaar

Abstract

A full overview of scheduled events for NeurIPS 2020 can be found here.

Nick Maxfield

From 2020 to 2022, Nick oversaw the van der Schaar Lab’s communications, including media relations, content creation, and maintenance of the lab’s online presence.