Die DPG-Frühjahrstagung in Dresden musste abgesagt werden! Lesen Sie mehr ...
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
MM: Fachverband Metall- und Materialphysik
MM 26: Topical Session: Data Driven Materials Science - Machine Learning for Materials Properties
MM 26.3: Vortrag
Dienstag, 17. März 2020, 14:45–15:00, BAR 205
Machine learning modeling of magnetic ground state and Curie temperature — •Teng Long, Nuno Fortunato, Yixuan Zhang, Oliver Gutfleisch, and Hongbin Zhang — Institute of Materials Science, Technical University of Darmstadt, Darmstadt 64287, Germany
Magnetic materials have a plethora of applications ranging from information and communication technologies to energy harvesting and conversion. However, their functionalities are often limited by the magnetic ordering temperature. In this work, we performed machine learning on the magnetic ground state and the Curie temperature (Tc), with generic chemical and crystal structural descriptors. Using 2805 known intermetallic compounds, a random forest model is trained to classify ferromagnetic and antiferromagnetic compounds and to predict the Tc of the ferromagnets, with only 15 and 23 descriptors used, respectively. The resulting accuracy is about 86% for classification and 92% for regression (with a mean absolute error (MAE) of 55K). We found that composition based features are sufficient for both classification and regression, whereas structural descriptors improve the performance. Using the trained model, we predicted the magnetic ordering and Tc for the intermetallic magnetic materials in the Materials Project, with a MAE of 73K in comparison to the experimental reported Tc that has been collected by us. This work paves the way to accelerate the discovery of new ferromagnetic compounds for technological applications.