Talk of Dr Olivier Bachem

Event details
Date | 03.10.2019 |
Hour | 11:15 › 12:15 |
Speaker | Dr Olivier Bachem |
Location | |
Category | Conferences - Seminars |
Title:
Google Research Football: A Novel Reinforcement Learning Environment
Abstract:
Recent progress in the field of reinforcement learning has been accelerated by virtual learning environments such as video games, where novel algorithms and ideas can be quickly tested in a safe and reproducible manner. We introduce the Google Research Football Environment, a new reinforcement learning environment where agents are trained to play football in an advanced, physics-based 3D simulator. The resulting environment is challenging, easy to use and customize, and it is available under a permissive open-source license. In addition, it provides support for multiplayer and multi-agent experiments. We propose three full-game scenarios of varying difficulty with the Football Benchmarks and report baseline results for three commonly used reinforcement algorithms (IMPALA, PPO, and Ape-X DQN). We also provide a diverse set of simpler scenarios with the Football Academy and showcase several promising research directions.
Bio:
Olivier Bachem is a research scientist in the Google Brain Team interested in fundamental problems in machine learning and artificial intelligence. He received his PhD from ETH Zurich where he was supervised by Andreas Krause in the Learning & Adaptive Systems group. In his dissertation, he investigated coresets - small summaries of large data sets with theoretical guarantees - and other sampling methods for large-scale clustering. He also held a Google PhD Fellowship in Machine Learning and was an Associated Fellow at the Max Planck ETH Center for Learning Systems. Before that, he obtained a bachelor’s degree in economics (University of St. Gallen), a master’s degree in quantitative finance (ETH Zurich & University of Zurich) as well as a master’s degree in statistics (ETH Zurich) where he was awarded an ETH medal for his master thesis.
Google Research Football: A Novel Reinforcement Learning Environment
Abstract:
Recent progress in the field of reinforcement learning has been accelerated by virtual learning environments such as video games, where novel algorithms and ideas can be quickly tested in a safe and reproducible manner. We introduce the Google Research Football Environment, a new reinforcement learning environment where agents are trained to play football in an advanced, physics-based 3D simulator. The resulting environment is challenging, easy to use and customize, and it is available under a permissive open-source license. In addition, it provides support for multiplayer and multi-agent experiments. We propose three full-game scenarios of varying difficulty with the Football Benchmarks and report baseline results for three commonly used reinforcement algorithms (IMPALA, PPO, and Ape-X DQN). We also provide a diverse set of simpler scenarios with the Football Academy and showcase several promising research directions.
Bio:
Olivier Bachem is a research scientist in the Google Brain Team interested in fundamental problems in machine learning and artificial intelligence. He received his PhD from ETH Zurich where he was supervised by Andreas Krause in the Learning & Adaptive Systems group. In his dissertation, he investigated coresets - small summaries of large data sets with theoretical guarantees - and other sampling methods for large-scale clustering. He also held a Google PhD Fellowship in Machine Learning and was an Associated Fellow at the Max Planck ETH Center for Learning Systems. Before that, he obtained a bachelor’s degree in economics (University of St. Gallen), a master’s degree in quantitative finance (ETH Zurich & University of Zurich) as well as a master’s degree in statistics (ETH Zurich) where he was awarded an ETH medal for his master thesis.
Practical information
- Expert
- Registration required
Organizer
- Professor Volkan Cevher
Contact
- Gosia Baltaian