Mila members to stand out at ICML 2021

Next month, Mila researchers will showcase their work in full force at the 38th International Conference on Machine Learning (ICML). This annual event brings together some of the brightest minds of the machine learning research community and will run virtually from July 18 to 24.

Of the 5,513 submissions this year, 1,184 papers were accepted (21.5%); 30 of those papers were co-authored by Mila researchers. 

The work being presented by Mila researchers covers a wide range of topics and demonstrates the far-reaching potential applications of machine learning in areas such as health, chemical physics, and even the airline industry. 

In “Structured Convolutional Kernel Networks for Airline Crew Scheduling,” Mila core academic member Simon Lacoste-Julien and colleagues introduce a novel deep structured predictor: structured convolutional kernel networks (Struct-CKN). The initial solutions obtained using Struct-CKN on a flight-connection dataset revealed that it can be further refined by an airline crew scheduling solver. 

Core faculty member Jian Tang and team designed a unified framework, Local-instance and Global-semantic Learning (GraphLoG), for self-supervised whole-graph representation learning, proposing hierarchical prototypes to capture global semantic clusters. The verified experiments on chemical and biological benchmark data sets demonstrate the method’s effectiveness, with potential implications for many tasks such as molecule properties prediction in drug and material discovery. 

On the other end of the spectrum, in “Continuous Coordination As a Realistic Scenario for Lifelong Learning,” Mila University professors Sarath Chandar and Aaron Courville, as well as research intern Hadi Nekoei and Master’s student Akilesh Badrinaaraayanan introduce Lifelong Hanabi, a continual multi-agent reinforcement learning testbed in which every task is coordinating with a partner that’s an expert player of Hanabi.

Below, accepted papers (spotlight and oral), workshops, and tutorials co-authored and co-organized by Mila members:

Tutorials

Continual learning with deep architectures
Vincenzo Lomonaco, Irina Rish
https://icml.cc/Conferences/2021/Schedule?showEvent=10833

Random Matrix Theory and ML (RMT+ML)
Fabian Pedregosa, Jeffrey Pennington, Thomas Trogdon, Courtney Paquette
https://icml.cc/Conferences/2021/Schedule?showEvent=10840

Oral Presentations

Directional Graph Networks
Dominique Beaini, Saro Passaro, Vincent Létourneau, Will Hamilton, Gabriele Corso, Pietro Lió
https://arxiv.org/abs/2010.02863

Can Subnetwork Structure Be the Key to Out-of-Distribution Generalization?
Dinghuai Zhang, Kartik Ahuja, Yilun Xu, Yisen Wang, Aaron Courville
https://arxiv.org/pdf/2106.02890.pdf

Out-of-Distribution Generalization via Risk Extrapolation (REx)
David Krueger, Ethan Caballero, Joern-Henrik Jacobsen, Amy Zhang, Jonathan Binas, Dinghuai Zhang, Rémi Le Priol, Aaron Courville
https://arxiv.org/abs/2003.00688

Learning Gradient Fields for Molecular Conformation Generation
Chence Shi, Shitong Luo, Minkai Xu, Jian Tang
https://arxiv.org/abs/2105.03902

On Disentangled Representations Learned from Correlated Data
Frederik Träuble, Elliot Creager, Niki Kilbertus, Francesco Locatello, Andrea Dittadi, Anirudh Goyal, Bernhard Schölkopf, Stefan Bauer
https://arxiv.org/abs/2006.07886

RNN with Particle Flow for Probabilistic Spatio-temporal Forecasting
Soumyasundar Pal, Liheng Ma, Yingxue Zhang, Mark Coates 
https://arxiv.org/abs/2106.06064

Spotlight Presentations

Aggregating from Multiple Target-Shifted Sources
Changjian Shui, Zijian Li, Jiaqi Li, Christian Gagné, Charles X. Ling, Boyu Wang
https://arxiv.org/abs/2105.04051

Continuous Coordination As a Realistic Scenario For Lifelong Learning
Hadi Nekoei, Akilesh Badrinaaraayanan, Aaron Courville, Sarath Chandar
https://arxiv.org/abs/2103.03216

Diffusion Earth Mover’s Distance and Distribution Embeddings
Alexander Tong, Guillaume Huguet, Amine Natik, Kincaid MacDonald, Manik Kuchroo, Ronald Coifman, Guy Wolf, Smita Krishnaswamy
https://arxiv.org/abs/2102.12833

Learning a Universal Template for Few-shot Dataset Generalization
Eleni Triantafillou, Hugo Larochelle, Richard Zemel, Vincent Dumoulin
https://arxiv.org/abs/2105.07029

Self-supervised Graph-level Representation Learning with Local and Global Structure
Minghao Xu, Hang Wang, Bingbing Ni, Hongyu Guo, Jian Tang
https://arxiv.org/pdf/2106.04113.pdf

Affine Invariant Analysis of Frank-Wolfe on Strongly Convex Sets
Thomas Kerdreux, Lewis Liu, Simon Lacoste-Julien, Damien Scieur
https://arxiv.org/abs/2011.03351

Infinite-Dimensional Optimization for Zero-Sum Games via Variational Transport
Lewis Liu, Yufeng Zhang, Zhuoran Yang, Reza Babanezhad, Zhaoran Wang
https://opt-ml.org/papers/2020/paper_92.pdf

Simultaneous Similarity-based Self-Distillation for Deep Metric Learning
Karsten Roth, Timo Milbich, Bjorn Ommer, Joseph Paul Cohen, Marzyeh Ghassemi
https://arxiv.org/abs/2009.08348

Structured Convolutional Kernel Networks for Airline Crew Scheduling
Yassine Yaakoubi, François Soumis, Simon Lacoste-Julien 
https://arxiv.org/abs/2105.11646

An End-to-End Framework for Molecular Conformation Generation via Bilevel Programming
Minkai Xu, Wujie Wang, Shitong Luo, Chence Shi, Yoshua Bengio, Rafael Gomez-Bombarelli, Jian Tang
https://arxiv.org/abs/2105.07246

Catastrophic Fisher Explosion: Early Phase Fisher Matrix Impacts Generalization
Stanislaw Jastrzebski, Devansh Arpit, Oliver Astrand, Giancarlo Kerg, Huan Wang, Caiming Xiong, Richard Socher, Kyunghyun Cho, Krzysztof J Geras
https://arxiv.org/abs/2012.14193

Preferential Temporal Difference Learning
Nishanth Anand, Doina Precup
https://arxiv.org/abs/2106.06508

Exploration in Approximate Hyper-State Space for Meta Reinforcement Learning
Luisa Zintgraf, Leo Feng, Cong Lu, Maximilian Igl, Kristian Hartikainen, Katja Hofmann, Shimon Whiteson
https://arxiv.org/abs/2010.01062

Equivariant Networks for Pixelized Spheres
Mehran Shakerinava, Siamak Ravanbakhsh

OptiDICE: Offline Policy Optimization via Stationary Distribution Correction Estimation
Jongmin Lee, Wonseok Jeon, Byung-Jun Lee, Joelle Pineau, Kee-Eung Kim

Beyond Variance Reduction: Understanding the True Impact of Baselines on Policy Optimization
Wesley Chung, Valentin Thomas, Marlos C. Machado, Nicolas Le Roux 
https://arxiv.org/abs/2008.13773

Randomized Exploration in Reinforcement Learning with General Value Function Approximation
Haque Ishfaq, Qiwen Cui, Alex Ayoub, Viet Nguyen, Zhuoran Yang, Zhaoran Wang, Doina Precup, Lin Yang

Locally Persistent Exploration in Continuous Control Tasks with Sparse Rewards
Susan Amin, Maziar Gomrokchi, Hossein Aboutalebi, Harsh Satija, Doina Precup 
https://arxiv.org/abs/2012.13658

Non-Autoregressive Electron Redistribution Modeling for Reaction Prediction
Hangrui Bi, Hengyi Wang, Chence Shi, Connor Coley, Jian Tang, Hongyu Guo
https://icml.cc/Conferences/2021/ScheduleMultitrack?event=9427

Trajectory Diversity for Zero-Shot Coordination
Andrei Lupu, Brandon Cui, Hengyuan Hu, Jakob Foerster
https://icml.cc/Conferences/2021/ScheduleMultitrack?event=9434

Multi-Task Reinforcement Learning with Context-based Representations
Shagun Sodhani, Amy Zhang, Joelle Pineau
https://arxiv.org/pdf/2102.06177.pdf

Robust Representation Learning via Perceptual Similarity Metrics
Saeid A Taghanaki, Kristy Choi, Amir Hosein Khasahmadi, Anirudh Goyal 

Vector Quantized Models for Planning
Sherjil Ozair, Yazhe Li, Ali Razavi, Ioannis Antonoglou, Aäron van den Oord, Oriol Vinyals 
https://arxiv.org/abs/2106.04615

Efficient Deviation Types and Learning for Hindsight Rationality in Extensive-Form Games
Dustin Morrill, Ryan D’Orazio, Marc Lanctot, James Wright, Michael Bowling, Amy Greenwald 
https://arxiv.org/abs/2102.06973

A Deep Reinforcement Learning Approach to Marginalized Importance Sampling with the Successor Representation
Scott Fujimoto, David Meger, Doina Precup

Workshops

Tackling Climate Change with Machine Learning
Hari Prasanna Das, Meareg Hailemariam, Katarzyna Tokarska, Maria João Sousa, David Rolnick, Xiaoxiang Zhu, Yoshua Bengio
Workshop Home Page

Unsupervised Reinforcement Learning
Feryal Behbahani, Joelle Pineau, Lerrel Pinto, Roberta Raileanu, Aravind Srinivas, Denis Yarats, Amy Zhang
Workshop Home Page

INNF+: Invertible Neural Networks, Normalizing Flows, and Explicit Likelihood Models
Chin-Wei Huang, David Krueger, Rianne Van den Berg, George Papamakarios, Tian Qi Chen, Danilo J. Rezende
Workshop Home Page

Theory and Foundation of Continual Learning
Thang Doan, Bogdan Mazoure, Amal Rannen Triki, Rahaf Aljundi, Vincenzo Lomonaco, Xu He, Arslan Chaudhry
Workshop Home Page

Women in Machine Learning virtual Un-workshop
Wenshuo Guo, Amy Zhang, Nezihe Merve Gürel, Audrey Durand, Ehi Nosakhare, Christina Papadimitriou, Sarah Poole, Tatjana Chavdarova
Workshop Home Page