Tensile2d

The dataset Tensile2d is available in PLAID format here. PLAID (Physics Informed AI Datamodel) is a library proposing an implementation for a datamodel tailored for AI and ML learning of physics problems. As for MMGP, PLAID has been developped at SafranTech. The code is hosted on Safran Gitlab, and a documentation is available on readthedocs.

The dataset Tensile2d is used as numerical experiment in the MMGP paper. The machine learning problem is presented in the PLAID documention, under data challenges.

Finally, the files required to run the MMGP method on the dataset are available in the MMGP’s repository example folder. In particular, refer to the example readme. We mention here that the example is set with the scikit-learn backend for ease of use, although the GPy backend was used in the MMGP paper.