Challenges of real-world reinforcement learning: definitions, benchmarks and analysis
| Autor: | Daniel J. Mankowitz, Cosmin Paduraru, Gabriel Dulac-Arnold, Nir Levine, Jerry Li, Todd Hester, Sven Gowal |
|---|---|
| Rok vydání: | 2021 |
| Předmět: |
Computer science
Suite Control (management) Context (language use) 02 engineering and technology Data science Artificial Intelligence 020204 information systems 0202 electrical engineering electronic engineering information engineering Benchmark (computing) Leverage (statistics) Reinforcement learning 020201 artificial intelligence & image processing Markov decision process Set (psychology) Software |
| Zdroj: | Machine Learning. 110:2419-2468 |
| ISSN: | 1573-0565 0885-6125 |
| Popis: | Reinforcement learning (RL) has proven its worth in a series of artificial domains, and is beginning to show some successes in real-world scenarios. However, much of the research advances in RL are hard to leverage in real-world systems due to a series of assumptions that are rarely satisfied in practice. In this work, we identify and formalize a series of independent challenges that embody the difficulties that must be addressed for RL to be commonly deployed in real-world systems. For each challenge, we define it formally in the context of a Markov Decision Process, analyze the effects of the challenge on state-of-the-art learning algorithms, and present some existing attempts at tackling it. We believe that an approach that addresses our set of proposed challenges would be readily deployable in a large number of real world problems. Our proposed challenges are implemented in a suite of continuous control environments called realworldrl-suite which we propose an as an open-source benchmark. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full text from SpringerLink