Review of the Noisy Trust Region Method
-
Updated
May 9, 2023 - Jupyter Notebook
Review of the Noisy Trust Region Method
SNR-driven adaptive step-size control for CMA-ES under noisy black-box optimization. A transparent proxy layer that adjusts sigma using signal-to-noise ratio estimates, with full experimental infrastructure and sensitivity analysis.
An Implementation of the N-Tuple Bandits Evolutionary Algorithm.
To associate your repository with the noisy-optimization topic, visit your repo's landing page and select "manage topics."