Summary of Learning Robust Reward Machines From Noisy Labels, by Roko Parac et al.
Learning Robust Reward Machines from Noisy Labels
by Roko Parac, Lorenzo Nodari, Leo Ardon, Daniel Furelos-Blanco, Federico Cerutti, Alessandra Russo
First submitted to arxiv on: 27 Aug 2024
Categories
- Main: Artificial Intelligence (cs.AI)
- 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 This paper proposes PROB-IRM, a novel approach for learning robust reward machines (RMs) in reinforcement learning (RL). RMs decompose tasks into subtasks, enabling RL agents to exploit their knowledge. PROB-IRM utilizes an inductive logic programming framework to learn RMs from noisy execution traces, leveraging Bayesian posterior degree of beliefs to ensure robustness. The method interleaves RM learning and policy learning, using probabilistic reward shaping to speed up training. Experimental results demonstrate that PROB-IRM can successfully train RL agents to solve tasks despite learning imperfect RMs from noisy data. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Imagine a robot trying to learn how to do a task by observing someone else doing it. But the person is making mistakes, and the robot needs to figure out what’s going wrong. This paper shows how to teach robots (or “agents”) to learn from imperfect examples and use that knowledge to do better in the future. The method, called PROB-IRM, helps agents understand what they’re supposed to be doing and how to improve by looking at the mistakes they make. |
Keywords
» Artificial intelligence » Reinforcement learning
