This paper proposes a supervised classification and regression approach for superconductive materials using DeepSet technology. The method successfully predicts both whether a material is superconducting and its critical temperature (Tc). Three materials identified by the neural network were experimentally characterized, confirming superconductivity in a synthetic analogue of michenerite (PdBiTe) and, for the first time, in monchetundraite (Pd₂NiTe₃), with Tc values matching predictions. This is the first superconducting material identified by AI.
Publisher
npj Computational Materials
Published On
May 02, 2023
Authors
Claudio Pereti, Kevin Bernot, Thierry Guizouarn, František Laufek, Anna Vymazalová, Luca Bindi, Roberta Sessoli, Duccio Fanelli
Tags
superconductivity
machine learning
DeepSet technology
material science
critical temperature
Related Publications
Explore these studies to deepen your understanding of the subject.
