c-bata

See https://github.com/optuna/optuna/pull/1767

Goptuna cannot track the benchmark performance of pruning algorithms. Add kurobako problem for pruning features.

See https://github.com/optuna/optuna/pull/1228

Gaussian process surrogate models for the CMA-ES

2

http://artax.karlin.mff.cuni.cz/~bajel3am/gpcomparison2017/

Adversarially Robust Optimization with Gaussian Processes

1

https://papers.nips.cc/paper/7818-adversarially-robust-optimization-with-gaussian-processes.pdf

Zhongxiang Dai, Haibin Yu, Bryan Kian Hsiang Low, and Patrick Jaillet. Bayesian optimization meets Bayesian optimal stopping. In Proc. ICML, pages 1496–1506, 2019 * Paper: http://proceedings.mlr.press/v97/dai19a.html * Code: https://github.com/daizhongxiang/Bayesian-Optimization-Meets-Bayesian-Optimal-Stopping

CI is failed at #109 due to the changes of black. I reformatted the code with black 22.3.0, the latest version to resolve CI issue.

Run tests with Python 3.10

1

This PR aims to support Python 3.10. Note that `3.10` must be quoted, otherwise tests are executed with Python 3.1.

## TODO * [ ] https://github.com/CyberAgent/cmaes/blob/main/cmaes/cma.py

This is a followup task for https://github.com/CyberAgent/cmaes/pull/87