The (1+1)-CMA-ES with margin is an extension of CMA-ES with margin [1] with the elite-preserving strategy. The (1+1)-CMA-ES with margin works effectively on binary and integer black-box optimization problems as well as mixed-integer black-box optimization problems.
sample_script_mixed.ipynb- Sample code for mixed-integer black-box optimization problems
sample_script_integer.ipynb- Sample code for integer black-box optimization problems
sample_script_binary.ipynb- Sample code for binary black-box optimization problems
demonstration_of_mean_vector_discretization.ipynb- Experiment about mean vector discretization (see Sec.3.2 and Fig.1 in our paper for GECCO 2023)
We tested this source code with the following environment.
numpy==1.24.3
scipy==1.10.1
Yohei Watanabe, Kento Uchida, Ryoki Hamano, Shota Saito, Masahiro Nomura, and Shinichi Shirakawa: (1+1)-CMA-ES with Margin for Discrete and Mixed-Integer Problems, Genetic and Evolutionary Computation Conference Companion (GECCO 2023), Lisbon, Portugal (hybrid), July 15-19, 2023. [arXiv]
[1] Ryoki Hamano, Shota Saito, Masahiro Nomura, and Shinichi Shirakawa, CMA-ES with Margin: Lower-Bounding Marginal Probability for Mixed-Integer Black-Box Optimization, In Genetic and Evolutionary Computation Conference (GECCO ’22), July 9–13, 2022, Boston, MA, USA. ACM, New York, NY, USA, 9 pages. https://doi.org/10.1145/3512290.3528827 [arXiv]
Our code is based on the source code of CMA-ES with margin [link]