ML@GT at ICML 2019
Georgia Tech Researchers Present at Global Machine Learning Conference
Georgia Tech researchers will present 18 research papers at this year’s event. The papers touch on a variety of aspects of machine learning including blended unconditional gradients, clustering with fairness constraints, and observational agents.
School of Interactive Computing assistant professor, Byron Boots is a 2019 area chair. Boots is also the co-organizer of the Real-World Sequential Decision Making: Reinforcement Learning and Beyond workshop and a guest speaker at the Generative Modeling and Model-Based Reasoning for Robotics and AI.
“ICML is globally renowned as one of the best conferences for machine learning research. Year after year, cutting edge research is presented and published and it’s a sign of ML@GT’s strength that Georgia Tech is consistently a top contributor in the accepted papers.” Justin Romberg, School of Electrical and Computer Engineering Schlumberger Professor and associate director of the Machine Learning Center at Georgia Tech (ML@GT).
Hosted June 9 through 15 at the Long Beach Convention and Entertainment Center, ICML is one of the fastest growing conferences in the world. It will bring together over 8,000 participants including entrepreneurs, engineers, graduate students, postdocs, and academic and industrial researchers.
Along with Georgia Tech papers, other accepted papers will include work in closely related fields like statistics, data science, and artificial intelligence, and important application areas like speech recognition, robotics, and machine vision.
RESEARCH HIGHLIGHTS
We hate to brag (do we?), but our students and faculty are producing some pretty cool research. Here are recaps on just a few of our papers at ICML 2019.
Imitating Latent Policies from Observation
By Ashley Edwards, Himanshu Sahni, Yannick Schroecker, Charles Isbell
Researchers explore a new approach that uses imitation learning from observation and video data. This new way of thinking could eventually teach agents how to do tasks like make a sandwich, play a videogame, or even drive a car, all from watching videos. Attend their presentation on Wednesday, June 12 at 11:25 AM in Hall B or at poster #33 during the poster session from 6:30-9:00 PM.
New Machine Learning Algorithms Keep Group Data Diverse
By Matthäus Kleindessner, Samira Samadi, Pranjal Awasthi, Jamie Morgenstern
Georgia Tech researchers have created machine learning (ML) algorithms to ensure grouped data is fairly represented.
This is the first example of incorporating fairness into the popular spectral clustering technique for partitioning graph data, according to researchers.
See their presentation on Thursday, June 13 at 12:10 PM in Room 103 or their poster session from 6:30-9 PM at poster #195.
Blended Conditional Gradients: The Unconditioning of Conditional Gradients
By Gábor Braun, Sebastian Pokutta, Dan Tu, Stephen Wright
ML@GT Associate Director, Sebastian Pokutta, gives an informal summary of his recent paper that shows how mixing the Frank-Wolfe and Gradient Descent creates a new, fast, projection-free algorithm for constrained smooth convex minimization.
Catch the oral presentation at 2:20 in Room 103 or the poster session from 6:30-9 PM at poster #191 on Tuesday, June 11.
Georgia Tech Papers
Generative Adversarial User Model for Reinforcement Learning Based Recommendation System Read
Xinshi Chen, Shuang Li, Hui Li, Shaohua Jiang, Yuan Qi, Le Song
Particle Flow Bayes' Rule
Xinshi Chen, Hanjun Dai, Le Song
Differentiable Decoding of Sets of Sequences for Neural Sequence Models
Ashwin Kalyan, Peter Anderson, Stefan Lee, Dhruv Batra
Probabilistic Neural-symbolic Models for Interpretable Visual Question Answering
Ramakrishna Vedantam, Karan Desai, Stefan Lee, Marcus Rohrbach, Dhruv Batra, Devi Parikh
TarMAC: Targeted Multi-Agent Communication
Abhishek Das, Théophile Gervet, Joshua Romoff, Dhruv Batra, Devi Parikh, Michael Rabbat, Joelle Pineau
Counterfactual Visual Explanations
Yash Goyal, Ziyan Wu, Jan Ernst, Dhruv Batra, Devi Parikh, Stefan Lee
Predictor-Corrector Policy Optimization
Ching-An Cheng, Xinyan Yan, Nathan Ratliff, and Byron Boots
Kernel Mean Matching for Content Addressability of GANs
Wittawat Jitkrittum, Patsorn Sangkloy, Muhammad Waleed Gondal, Amit Raj, James Hays, Bernhard Schölkopf
Latent Space Regularization for Explicit Control of Musical Attributes
Kumar Ashis Pati and Alexander Lerch
Provably Efficient Imitation Learning from Observation Alone
Wen Sun, Anirudh Vemula, Byron Boots, and Drew Bagnell
Competing Against Equilibria in Zero-Sum Games with Evolving Payoffs
Adrian Rivera Cardoso, Jacob Abernethy, He Wang and Huan Xu
Fair k-Center Clustering for Data Summarization
Matthäus Kleindessner, Pranjal Awasthi, Jamie Morgenstern
Learning Novel Policies for Tasks
Yunbo Zhang, Wenhao Yu and Greg Turk
Active Embedding Search via Noisy Paired Comparisons
Gregory Canal, Andy Massimino, Mark Davenport, Christopher Rozell
Finite-Time Analysis of Distributed TD(0) with Linear Function Approximation on Multi-Agent Reinforcement Learning
Thinh T. Doan, Siva Theja Maguluri, Justin Romberg
Towards Understanding the Importance of Noise in Training Neural Networks
Mo Zhou, Tianyi Liu, Yan Li, Dachao Lin, Enlu Zhou, and Tuo Zhao
On Scalable and Efficient Computation of Large Scale Optimal Transport
Yujia Xie, Minshuo Chen, Haoming Jiang, Tuo Zhao, and Hongyuan Zha
A Comparison of Music Input Domains for Self-Supervised Feature Learning
Siddharth Gururani, Alexander Lerch and Mason Bretan
LIVE ICML PHOTO GALLERY
ABOUT ML@GT
WHO WE ARE
AREAS OF EXPERTISE
- Computer Vision
- Natural Language Processing
- Robotics
- Deep Learning
- Game Theory
- Neuro Computing
- Ethics and Fairness
- Artificial Intelligence
- Internet of Things
- Machine Learning Theory
- Systems for Machine Learning
- Bioinformatics
- Computational Finance
- Health Systems
- Information Security
- Logistics and Manufacturing