Ioannis Mitliagkas

verfied profile

∙ 0 followers
The University of Texas at Austin ∙ Université de Montréal ∙ Stanford University

I work on topics in optimization, dynamics and learning, with a focus on modern machine learning. I have done work in the intersection of systems and theory. Some recent topics:

- Min-max optimization and the dynamics of games

- Generalization and domain adaptation

- Optimization for deep learning

- Statistical learning and inference

Adversarial score matching and improved sampling for image generation

Denoising score matching with Annealed Langevin Sampling (DSM-ALS) is a ...
09/11/2020 ∙ by Alexia Jolicoeur-Martineau, et al. ∙ 10 ∙ share

read it
Stochastic Hamiltonian Gradient Methods for Smooth Games

The success of adversarial formulations in machine learning has brought ...
07/08/2020 ∙ by Nicolas Loizou, et al. ∙ 45 ∙ share

read it
Accelerating Smooth Games by Manipulating Spectral Shapes

We use matrix iteration theory to characterize acceleration in smooth ga...
01/02/2020 ∙ by Waïss Azizian, et al. ∙ 16 ∙ share

read it
Adversarial target-invariant representation learning for domain generalization

In many applications of machine learning, the training and test set data...
11/03/2019 ∙ by Isabela Albuquerque, et al. ∙ 0 ∙ share

read it
Connections between Support Vector Machines, Wasserstein distance and gradient-penalty GANs

We generalize the concept of maximum-margin classifiers (MMCs) to arbitr...
10/15/2019 ∙ by Alexia Jolicoeur-Martineau, et al. ∙ 48 ∙ share

read it
Lower Bounds and Conditioning of Differentiable Games

Many recent machine learning tools rely on differentiable game formulati...
06/17/2019 ∙ by Adam Ibrahim, et al. ∙ 0 ∙ share

read it
A Tight and Unified Analysis of Extragradient for a Whole Spectrum of Differentiable Games

We consider differentiable games: multi-objective minimization problems,...
06/13/2019 ∙ by Waïss Azizian, et al. ∙ 0 ∙ share

read it
Reducing the variance in online optimization by transporting past gradients

Most stochastic optimization methods use gradients once before discardin...
06/08/2019 ∙ by Sébastien M. R. Arnold, et al. ∙ 0 ∙ share

read it
State-Reification Networks: Improving Generalization by Modeling the Distribution of Hidden Representations

Machine learning promises methods that generalize well from finite label...
05/26/2019 ∙ by Alex Lamb, et al. ∙ 0 ∙ share

read it
SysML: The New Frontier of Machine Learning Systems

Machine learning (ML) techniques are enjoying rapidly increasing adoptio...
03/29/2019 ∙ by Alexander Ratner, et al. ∙ 0 ∙ share

read it
Multi-objective training of Generative Adversarial Networks with multiple discriminators

Recent literature has demonstrated promising results for training Genera...
01/24/2019 ∙ by Isabela Albuquerque, et al. ∙ 0 ∙ share

read it
A Modern Take on the Bias-Variance Tradeoff in Neural Networks

We revisit the bias-variance tradeoff for neural networks in light of mo...
10/19/2018 ∙ by Brady Neal, et al. ∙ 20 ∙ share

read it
h-detach: Modifying the LSTM Gradient Towards Better Optimization

Recurrent neural networks are known for their notorious exploding and va...
10/06/2018 ∙ by Devansh Arpit, et al. ∙ 0 ∙ share

read it
Negative Momentum for Improved Game Dynamics

Games generalize the optimization paradigm by introducing different obje...
07/12/2018 ∙ by Gauthier Gidel, et al. ∙ 10 ∙ share

read it
Fortified Networks: Improving the Robustness of Deep Networks by Modeling the Manifold of Hidden Representations

Deep networks have achieved impressive results across a variety of impor...
04/07/2018 ∙ by Alex Lamb, et al. ∙ 0 ∙ share

read it
Deep Learning at 15PF: Supervised and Semi-Supervised Classification for Scientific Data

This paper presents the first, 15-PetaFLOP Deep Learning system for solv...
08/17/2017 ∙ by Thorsten Kurth, et al. ∙ 0 ∙ share

read it
Improving Gibbs Sampler Scan Quality with DoGS

The pairwise influence matrix of Dobrushin has long been used as an anal...
07/18/2017 ∙ by Ioannis Mitliagkas, et al. ∙ 0 ∙ share

read it
Accelerated Stochastic Power Iteration

Principal component analysis (PCA) is one of the most powerful tools in ...
07/10/2017 ∙ by Christopher De Sa, et al. ∙ 0 ∙ share

read it
Learning Representations and Generative Models for 3D Point Clouds

Three-dimensional geometric data offer an excellent domain for studying ...
07/08/2017 ∙ by Panos Achlioptas, et al. ∙ 0 ∙ share

read it
YellowFin and the Art of Momentum Tuning

Hyperparameter tuning is one of the big costs of deep learning. State-of...
06/12/2017 ∙ by Jian Zhang, et al. ∙ 0 ∙ share

read it
Parallel SGD: When does averaging help?

Consider a number of workers running SGD independently on the same pool ...
06/23/2016 ∙ by Jian Zhang, et al. ∙ 0 ∙ share

read it
Scan Order in Gibbs Sampling: Models in Which it Matters and Bounds on How Much

Gibbs sampling is a Markov Chain Monte Carlo sampling technique that ite...
06/10/2016 ∙ by Bryan He, et al. ∙ 0 ∙ share

read it
Asynchrony begets Momentum, with an Application to Deep Learning

Asynchronous methods are widely used in deep learning, but have limited ...
05/31/2016 ∙ by Ioannis Mitliagkas, et al. ∙ 0 ∙ share

read it
Memory Limited, Streaming PCA

We consider streaming, one-pass principal component analysis (PCA), in t...
06/28/2013 ∙ by Ioannis Mitliagkas, et al. ∙ 0 ∙ share

read it

Ioannis Mitliagkas

Featured Co-authors

Yoshua Bengio

Michael I. Jordan

Jian Zhang

Ce Zhang

Christopher Ré

Dawn Song

Peng Xu

Martin Jaggi

Prateek Jain

Simon Lacoste-Julien

Jing Li

Adversarial score matching and improved sampling for image generation

Stochastic Hamiltonian Gradient Methods for Smooth Games

Accelerating Smooth Games by Manipulating Spectral Shapes

Adversarial target-invariant representation learning for domain generalization

Connections between Support Vector Machines, Wasserstein distance and gradient-penalty GANs

Lower Bounds and Conditioning of Differentiable Games

A Tight and Unified Analysis of Extragradient for a Whole Spectrum of Differentiable Games

Reducing the variance in online optimization by transporting past gradients

State-Reification Networks: Improving Generalization by Modeling the Distribution of Hidden Representations

SysML: The New Frontier of Machine Learning Systems

Multi-objective training of Generative Adversarial Networks with multiple discriminators

A Modern Take on the Bias-Variance Tradeoff in Neural Networks

h-detach: Modifying the LSTM Gradient Towards Better Optimization

Negative Momentum for Improved Game Dynamics

Fortified Networks: Improving the Robustness of Deep Networks by Modeling the Manifold of Hidden Representations

Deep Learning at 15PF: Supervised and Semi-Supervised Classification for Scientific Data

Improving Gibbs Sampler Scan Quality with DoGS

Accelerated Stochastic Power Iteration

Learning Representations and Generative Models for 3D Point Clouds

YellowFin and the Art of Momentum Tuning

Parallel SGD: When does averaging help?

Scan Order in Gibbs Sampling: Models in Which it Matters and Bounds on How Much

Asynchrony begets Momentum, with an Application to Deep Learning

Memory Limited, Streaming PCA