Gashmard, H., Shakeripour, H.

Machine learning methods in studying of superconducting materials

(2022) 9 Feb. 29^th Symposium of Crystallography and Mineralogy of Iran.

Abstract

A comprehensive theory capable of predicting the critical temperature of all superconducting materials has not been proposed to date. Observing the critical temperature of a compound requires heavy costs in order to make a sample in the laboratory. In this study, we predicted the critical temperature of some superconducting material compounds using a machine learning model. Here we used a dataset containing 13562 superconducting compounds to calculate 132 properties for each compound. The value of R 2 was Obtained equal to 0.924 and the value of RMSE was Obtained equal to 7.3 Kelvin.

Keywords: Superconducting Materials, Machine Learning, XGBoost Mode