Summary of Instance Selection For Dynamic Algorithm Configuration with Reinforcement Learning: Improving Generalization, by Carolin Benjamins et al.
Instance Selection for Dynamic Algorithm Configuration with Reinforcement Learning: Improving Generalization
by Carolin Benjamins, Gjorgjina Cenikj, Ana Nikolikj, Aditya Mohan, Tome Eftimov, Marius Lindauer
First submitted to arxiv on: 18 Jul 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: None
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 proposes a novel approach to Dynamic Algorithm Configuration (DAC) by addressing the challenge of generalizing agents trained with Deep Reinforcement Learning (RL). The authors suggest that the limited generalization performance is due to bias in the training instances and propose a method to mitigate this by selecting a representative subset of training instances. They use meta-features, including time series features on trajectories of actions and rewards, to construct the subset. Empirical evaluations on Sigmoid and CMA-ES benchmarks from DACBench demonstrate the efficacy of instance selection in refining DAC policies for diverse instance spaces. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary DAC is a challenge where agents are trained with Deep Reinforcement Learning (RL) but struggle to generalize well across different instances. The paper suggests that this might be due to biased training and proposes a solution by selecting a smaller, more representative set of training instances. This can help improve the generalization performance of the agent. The authors use special features to choose which instances to keep and show that this approach works better than just using all the training data. |
Keywords
» Artificial intelligence » Generalization » Reinforcement learning » Sigmoid » Time series
