MTL MLOpt

Montréal Machine Learning and Optimization (MTL MLOpt) is a group of researchers living and working in Montréal.

Our research loosely spans topics in machine learning and mathematical optimization. Many of our members are affiliated with the Mila, where we also held our physical meetings (in pre-apocalyptic times). The group includes researchers from the University of Montréal, McGill, Google Brain, Samsung SAIT AI Lab (SAIL) Montreal, Facebook AI Research Montréal (FAIR) and Microsoft Research Montréal.

We hold public and internal meetings. The public meetings are a seminar with guest lecturers from around the world. Our public meetings are open for everyone to attend. Our internal meetings typically comprise of a presentation of a member's latest work, followed by discussion which often leads to productive collaborations.

Join our mailing group to receive the latest public meeting announcements.
Find the videos from our seminar here.
Follow us on Twitter for more updates.

Organizers: Reyhane Askari Hemmat, Arinka Jancarik, Nicolas Le Roux, Gauthier Gidel, Ioannis Mitliagkas

Summer 2021

July 14th	Mark Sellke	A Universal Law of Robustness via Isoperimetry	[paper]
June 30th	Ludwig Schmidt	Evaluating Machine (Human) Accuracy and Robustness on ImageNet
June 16th	Jelena Diakonikolas	Structure in Min-Max Optimization (and How to Use It!)
June 2nd	Tong Zhang	Why Cosine Learning Rate Scheduler Works and How to Improve It

Spring 2021

May 5th	Vasilis Syrgkanis	Adversarial machine learning and instrumental variables for flexible causal modeling
April 21st	Lenka Zdeborova	Insights on Gradient-based algorithms in high-dimensional non-convex optimisation
April 7th	Aaron Defazio	Why are we still using SGD in the 21st century? Adventures in large-scale deep learning
March 24th	Sebastien Bubeck	A law of robustness for two-layers neural networks

Winter 2021

March 10th	Margarida Carvalho	Combinatorial optimization for games
February 24th	Anastasios Kyrillidis	Distributed learning of deep neural networks using independent subnet training
February 10th	Lorenzo Rosasco	An implicit tour of regularization
January 27th	Robert M. Gower	New Viewpoints, Variants and Convergence Theory for Stochastic Polyak Stepsizes
January 13th	Panayotis Mertikopoulos	Dynamics, (min-max) optimization, and games

Fall 2020

December 2nd	Nicolas Loizou	SGD for Modern Machine Learning: Practical Variants and Convergence Guarantees
November 4th	Ashia Wilson	Variational Perspectives on Machine Learning: Algorithms, Inference, and Fairness
October 7th	Karolina Dziugaite	Distribution-dependent generalization bounds for noisy, iterative learning algorithms
September 23rd	Aude Genevay	Learning with entropy-regularized optimal transport
September 9th	Geoffrey Negiar	Stochastic Frank-Wolfe for Constrained Finite-Sum Minimization

Summer 2020

August 26th	Hanie Sedghi	What is being transferred in transfer learning?
August 12th	Kamalika Chaudhuri	Challenges in Reliable Machine Learning
August 7th	Rachel Ward	Weighted Optimization: better generalization by smoother interpolation
July 31st	Costis Daskalakis	The Complexity of Min-Max Optimization
July 3rd	Rachael Tappenden	Accelerated Gradient Methods with Optimality Certificates	[paper]
June 5th	Francis Bach	On the effectiveness of Richardson Extrapolation in Machine Learning	[paper] [video]

Spring 2020

May 22nd	Tim Hoheisel	Cone-Convexity and Composite Functions	[paper]
May 8th	Peter Richtarik	On Second Order Methods and Randomness	[paper] [video]
May 1st	Adam Oberman	Accelerated stochastic gradient descent: convergence rate and empirical results	[paper]
April 17th	Mark Schmidt	Faster Algorithms for Deep Learning?	[video]

Winter 2020

March 27th, Chris Finlay, How to train your Neural ODE (paper)
March 13th, Ryan D'Orazio, Alternative Function Approximation Parameterizations for Solving Games (paper)
Feb. 28th, Mido Assran, On the convergence of Nesterov's Accelerated Gradient Method in stochastic settings
Feb. 14th, Damien Scieur, Acceleration through spectral density estimation V2 + Universal average-case optimality of Polyak momentum,

Fall 2019

Acceleration through spectral density estimation, Fabian Pedregosa, 2019/12/06
Adaptive regularization with inexact gradient for machine learning, Sanae Lotfi, 2019/11/22
Cross-over session with Mila's RL reading group, 2019/11/08
Stochastic Optimization, Nicolas Loizou, 2019/10/25
Path Length Bounds for Gradient Descent and Flow (paper), Aaditya Ramdas, 2019/10/11
Implicit regularization in deep learning: a view from function space, Aristide Baratin, 2019/10/11
The Bias-Variance tradeoff in neural networks (paper), Brady Neal, 2019/09/13

Summer 2019

Painless Stochastic Gradient: Interpolation, Line-Search, and Convergence Rates (paper), Sharan Vaswani, 2019/08/30
From maximum likelihood to classification accuracy, Nicolas Le Roux, 2019/08/16
The duality gap technique (Diakonikolas, Orecchia), presented by Damien Scieur, 2019/08/02
Extragradient and current questions on game optimization, Waïss Azizian, 2019/07/19
Lower bounds and Conditioning of Differentiable Games (paper), Adam Ibrahim, 2019/07/05
ODE for Nesterov acceleration stochastic case, Maxime Laborde, 2019/07/05
Vector field perspective on GANs, Gauthier Gidel, 2019/06/21
Methods for adaptive SGD, Mariana Oliveira Prazeres, 2019/06/21