Machine learning modeling of superconducting critical temperature

By Valentin G. Stanev, Corey Oses, A. Kusne, Efrain Rodriguez, J. Paglione, Stefano Curtarolo, Ichiro Takeuchi. npj Computational Materials, 2017

446 citations recorded by the source corpus.

Superconductivity has been the focus of enormous research effort since its discovery more than a century ago. Yet, some features of this unique phenomenon remain poorly understood; prime among these is the connection between superconductivity and chemical/structural properties of materials. To bridge the gap, several machine learning schemes are developed herein to model the critical temperatures (Tc) of the 12,000+ known superconductors available via the SuperCon database. Materials are first divided into two classes based on their Tc values, above and below 10 K, and a classification model predicting this label is trained. The model uses coarse-grained features based only on the chemical compositions. It shows strong predictive power, with out-of-sample accuracy of about 92%. Separate regression models are developed to predict the values of Tc for cuprate, iron-based, and low-Tc compounds. These models also demonstrate good performance, with learned predictors offering potential insights into the mechanisms behind superconductivity in different families of materials. To improve the accuracy and interpretability of these models, new features are incorporated using materials data from the AFLOW Online Repositories. Finally, the classification and regression models are combined into a single-integrated pipeline and employed to search the entire Inorganic Crystallographic Structure Database (ICSD) for potential new superconductors. We identify >30 non-cuprate and non-iron-based oxides as candidate materials.Superconductivity: machine learning predicts superconducting transition temperatureMachine learning schemes are developed to model the superconducting transition temperature of over 12,000 compounds with good accuracy. A team led by Valentin Stanev from the University of Maryland at College Park and including researchers from Duke University and NIST develops several machine learning schemes to model the critical temperature (Tc) of over 12,000 known superconductors and candidate materials. They first train a classification model based only on the chemical compositions to categorize the known superconductors according to whether their Tc is above or below 10 K. Then they develop regression models to predict the values of Tc for various compounds. The accuracy of these models is further improved by including data from the AFLOW Online Repositories. They combine the classification and regression models into a single-integrated pipeline to search the entire Inorganic Crystallographic Structure Database and predict more than 30 new candidate superconductors.

View Paper: https://www.semanticscholar.org/paper/d310118dcbd5c72eb8e009ace9449083dcd7f340

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Evaluating explorative prediction power of machine learning algorithms for materials discovery using k-fold forward cross-validation (2020, methodology): “Stanev [4] has been putting the effort in developing a classification model to classify the samples into two classes with above or below 10 K critical temperature and a regression model to predict the critical temperature using random forest (RF)[32] algorithm with Magpie descriptor [10].”
Recent advances and applications of deep learning methods in materials science (2021, methodology): “Such methods have been used for diverse DFT datasets mentioned above in Table 1 as well as experimental datasets such as SuperCon 155,156 for quick pre-screening applications.”
A perspective on conventional high-temperature superconductors at high pressure: Methods and materials (2019, background): “Finally, in the field of superconductivity there has been recent efforts to tackle this puzzling problem irrespective of the underlying mechanism of superconductivity [375, 376, 377], for instance to machine learn models of superconducting critical temperatures [378].”
Emerging materials intelligence ecosystems propelled by machine learning (2020, background): “…, vibrational spectra and entropy 61 , thermal conductivity 62 ) properties to more exotic or application-oriented cases, such as superconductivity 63 , ferroelectric Curie temperatures 64 , properties of topological insulators 65 , thermoelectric figure of merit 66 and hopping barriers in Li-ion…”
A data-driven statistical model for predicting the critical temperature of a superconductor (2018, methodology): “Valentin et al. (2017) also create a model to predict Tc.”
Big Data Creates New Opportunities for Materials Research: A Review on Methods and Applications of Machine Learning for Materials Design (2019, unknown): “…[7], catalytic activities [8], and acid dissociation constants [9], as well as the identiﬁcation of promising porous materials [10], polymer dielectrics [11], mixed oxide catalysts [12], organic light-emitting diode (OLED) materials [13], superconductors [14], and photovoltaic materials [15].”