Modelling magnetic material properties with uncertainty-aware neural networks

This paper demonstrates that incorporating uncertainty quantification into machine learning models significantly enhances the reliability and trustworthiness of magnetic material property predictions, from composition-based surrogate models to complex microstructure-sensitive tasks.

Original authors: Clemens Wager, Heisam Moustafa, Alexander Kovacs, Qais Ali, Harald Oezelt, Hayate Yamano, Masao Yano, Noritsugu Sakuma, Hyuga Hosoi, Akihito Kinoshita, Tetsuya Shoji, Akira Kato, Thomas Schrefl