Summary of An Investigation Of Offline Reinforcement Learning in Factorisable Action Spaces, by Alex Beeson et al.
An Investigation of Offline Reinforcement Learning in Factorisable Action Spaces
by Alex Beeson, David Ireland, Giovanni Montana
First submitted to arxiv on: 17 Nov 2024
Categories
- Main: Machine Learning (stat.ML)
- Secondary: Machine Learning (cs.LG)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The paper explores offline reinforcement learning (RL) in factorised discrete action spaces, a domain that has received limited attention despite its relevance to many real-world problems. The authors propose a factorised approach to mitigate overestimation bias in value estimates and present an empirical evaluation of several offline techniques adapted to this setting. They introduce a suite of datasets with varying quality and task complexity, making them available for public use alongside their code base. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Offline reinforcement learning is important because it can help solve problems where data collection is challenging or risky. The paper focuses on factorised action spaces, which are common in real-world applications but haven’t been studied much before. The authors develop a new approach and test several existing methods to see how well they work in this new setting. They also create their own datasets for testing and make them available online. |
Keywords
* Artificial intelligence * Attention * Reinforcement learning
