Position: Benchmarking is Limited in Reinforcement Learning Research
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910499656957952 |
|---|---|
| author | Jordan, Scott M. White, Adam da Silva, Bruno Castro White, Martha Thomas, Philip S. |
| author_facet | Jordan, Scott M. White, Adam da Silva, Bruno Castro White, Martha Thomas, Philip S. |
| contents | Novel reinforcement learning algorithms, or improvements on existing ones, are commonly justified by evaluating their performance on benchmark environments and are compared to an ever-changing set of standard algorithms. However, despite numerous calls for improvements, experimental practices continue to produce misleading or unsupported claims. One reason for the ongoing substandard practices is that conducting rigorous benchmarking experiments requires substantial computational time. This work investigates the sources of increased computation costs in rigorous experiment designs. We show that conducting rigorous performance benchmarks will likely have computational costs that are often prohibitive. As a result, we argue for using an additional experimentation paradigm to overcome the limitations of benchmarking. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2406_16241 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Position: Benchmarking is Limited in Reinforcement Learning Research Jordan, Scott M. White, Adam da Silva, Bruno Castro White, Martha Thomas, Philip S. Machine Learning Methodology Novel reinforcement learning algorithms, or improvements on existing ones, are commonly justified by evaluating their performance on benchmark environments and are compared to an ever-changing set of standard algorithms. However, despite numerous calls for improvements, experimental practices continue to produce misleading or unsupported claims. One reason for the ongoing substandard practices is that conducting rigorous benchmarking experiments requires substantial computational time. This work investigates the sources of increased computation costs in rigorous experiment designs. We show that conducting rigorous performance benchmarks will likely have computational costs that are often prohibitive. As a result, we argue for using an additional experimentation paradigm to overcome the limitations of benchmarking. |
| title | Position: Benchmarking is Limited in Reinforcement Learning Research |
| topic | Machine Learning Methodology |
| url | https://arxiv.org/abs/2406.16241 |