Conversation
Codecov Report❌ Patch coverage is
📢 Thoughts on this report? Let us know! |
nikivanstein
left a comment
nikivanstein
left a comment
There was a problem hiding this comment.
Please alter the text for BBO according to my comments. Maybe you can also ask @sofokliski8 to write a paragraph about the AutoML benchmark? Then we can incorporate that immediately.
Next step would be the main readme.md file.
Also I saw, that the AlphaEvolve benchmarks have a readme inside their folder, it might also be nice to add one to the BBO, AutoML and Kernel tuning one? We can also reference the papers (BBOB paper, SBOX-Cost and MA-BBOB paper) and IOH.
docs/benchmarks.rst
Outdated
| AutoML | ||
| ^^^^^^ | ||
|
|
||
| - Used to generate a set of machine learning pipelines using libraries `scikit-learn`. |
There was a problem hiding this comment.
We need to add more specifics here, emphasis the link to the OpenML platform etc. For this PR it is ok, we can ask Sofoklis to write this.
|
|
||
| - Contains a set of Black Box Optimisation problems and Multi-Affine Black Box Optimisation Problems from the ``ioh`` library. | ||
| - Includes a special benchmark ``SBOX-COST``, which is similar to BBOB but defined on the domain :math:`[-5, 5]^n`. | ||
|
|
There was a problem hiding this comment.
I would not call this whole section BBOB as this is the name of one specific benchmark.
I would say: "Black Box Optimization (BBO)"
Then stress the link with ioh and also point to the https://iohprofiler.github.io/IOHexp/ website.
It contains several benchmarks:
- BBOB (unconstrained noiseless functions)
- SBOX-COST (similar to BBOB but with box constraints)
- MA-BBOB (Many Affine BBOB).
- Real world applications from Photonics (The photonics problem also should be under this category).
3 participants
Refactored
iohblade/problems->iohblade/benchmarks.