yoshitomo-matsubara
commited on
Commit
•
2b660c6
1
Parent(s):
655ca5a
addd preprint info
Browse files
README.md
CHANGED
@@ -50,7 +50,7 @@ task_ids: []
|
|
50 |
|
51 |
- **Homepage:**
|
52 |
- **Repository:** https://github.com/omron-sinicx/srsd-benchmark
|
53 |
-
- **Paper:** Rethinking Symbolic Regression Datasets and Benchmarks for Scientific Discovery
|
54 |
- **Point of Contact:** [Yoshitaka Ushiku](mailto:yoshitaka.ushiku@sinicx.com)
|
55 |
|
56 |
### Dataset Summary
|
@@ -58,7 +58,7 @@ task_ids: []
|
|
58 |
Our SRSD (Feynman) datasets are designed to discuss the performance of Symbolic Regression for Scientific Discovery.
|
59 |
We carefully reviewed the properties of each formula and its variables in [the Feynman Symbolic Regression Database](https://space.mit.edu/home/tegmark/aifeynman.html) to design reasonably realistic sampling range of values so that our SRSD datasets can be used for evaluating the potential of SRSD such as whether or not an SR method con (re)discover physical laws from such datasets.
|
60 |
|
61 |
-
This is the Hard set of our SRSD-Feynman datasets, which consists of the following 50 different physics formulas:
|
62 |
|
63 |
| ID | Formula |
|
64 |
|-----------|---------------------------------------------------------------------------------------------|
|
@@ -203,11 +203,12 @@ MIT License
|
|
203 |
|
204 |
### Citation Information
|
205 |
|
|
|
206 |
```bibtex
|
207 |
@article{matsubara2022rethinking,
|
208 |
title={Rethinking Symbolic Regression Datasets and Benchmarks for Scientific Discovery},
|
209 |
author={Matsubara, Yoshitomo and Chiba, Naoya and Igarashi, Ryo and Tatsunori, Taniai and Ushiku, Yoshitaka},
|
210 |
-
journal={arXiv preprint arXiv
|
211 |
year={2022}
|
212 |
}
|
213 |
```
|
|
|
50 |
|
51 |
- **Homepage:**
|
52 |
- **Repository:** https://github.com/omron-sinicx/srsd-benchmark
|
53 |
+
- **Paper:** [Rethinking Symbolic Regression Datasets and Benchmarks for Scientific Discovery](https://arxiv.org/abs/2206.10540)
|
54 |
- **Point of Contact:** [Yoshitaka Ushiku](mailto:yoshitaka.ushiku@sinicx.com)
|
55 |
|
56 |
### Dataset Summary
|
|
|
58 |
Our SRSD (Feynman) datasets are designed to discuss the performance of Symbolic Regression for Scientific Discovery.
|
59 |
We carefully reviewed the properties of each formula and its variables in [the Feynman Symbolic Regression Database](https://space.mit.edu/home/tegmark/aifeynman.html) to design reasonably realistic sampling range of values so that our SRSD datasets can be used for evaluating the potential of SRSD such as whether or not an SR method con (re)discover physical laws from such datasets.
|
60 |
|
61 |
+
This is the ***Hard set*** of our SRSD-Feynman datasets, which consists of the following 50 different physics formulas:
|
62 |
|
63 |
| ID | Formula |
|
64 |
|-----------|---------------------------------------------------------------------------------------------|
|
|
|
203 |
|
204 |
### Citation Information
|
205 |
|
206 |
+
[[Preprint](https://arxiv.org/abs/2206.10540)]
|
207 |
```bibtex
|
208 |
@article{matsubara2022rethinking,
|
209 |
title={Rethinking Symbolic Regression Datasets and Benchmarks for Scientific Discovery},
|
210 |
author={Matsubara, Yoshitomo and Chiba, Naoya and Igarashi, Ryo and Tatsunori, Taniai and Ushiku, Yoshitaka},
|
211 |
+
journal={arXiv preprint arXiv:2206.10540},
|
212 |
year={2022}
|
213 |
}
|
214 |
```
|