/MFGPextreme

Multi-fidelity Bayesian experimental design framework

Primary LanguagePythonMIT LicenseMIT

MFGPextreme

Reference: Xianliang Gong, Yulin Pan, Multi-fidelity Bayesian experimental design to quantify extreme-event statistics, 2021.

Notes:

The code will be further extended to other quantity of interests, e.g. statistical expectation, exceeding probability, high-order moments. The single-fidelity version can be found at https://github.com/ablancha/gpsearch.

The data can be found at https://drive.google.com/drive/folders/1kQmqEu9utBe20KBOMsX_9aKhrAXerKwk?usp=sharing.

Requirements:

Scipy: 1.5.2
Numpy: 1.17.2
Scikit-learn 0.24.2
pyDOE: 0.3.8
Joblib: 0.13.2
Emukit: 0.4.8
KDEpy: 1.0.3

Questions/Remarks:

Questions can be forwarded to xlgong@umich.edu.