Machine learning modeling of superconducting critical temperature
Machine learning modeling of superconducting critical temperature
复制标题
DOI：
10.1038/s41524-018-0085-8
复制
发表时间：
2018-06-28
影响因子：
9.7
通讯作者：
Takeuchi, Ichiro
中科院分区：
文献类型：
--
作者：
Stanev, Valentin;Oses, Corey;Takeuchi, Ichiro
自一个多世纪前发现以来，超导一直是大量研究工作的焦点。然而，人们对这种独特现象的一些特征仍然知之甚少。其中最主要的是超导性和材料的化学/结构特性之间的联系。为了弥补这一差距，本文开发了几种机器学习方案，对 SuperCon 数据库中提供的 12,000 多种已知超导体的临界温度 (T-c) 进行建模。首先根据材料的 T-c 值（高于和低于 10 K）将材料分为两类，并训练预测该标签的分类模型。该模型使用仅基于化学成分的粗粒度特征。它显示出强大的预测能力，样本外准确率约为 92%。开发了单独的回归模型来预测铜酸盐、铁基和低 T-c 化合物的 T-c 值。这些模型还表现出了良好的性能，学习的预测器为不同材料家族的超导性背后的机制提供了潜在的见解。为了提高这些模型的准确性和可解释性，使用 AFLOW 在线存储库中的材料数据合并了新功能。最后，分类和回归模型被组合成一个单一的集成管道，并用于搜索整个无机晶体结构数据库（ICSD）以寻找潜在的新型超导体。我们确定了超过 30 种非铜酸盐和非铁基氧化物作为候选材料。
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 (T-c) of the 12,000+ known superconductors available via the SuperCon database. Materials are first divided into two classes based on their T-c 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 T-c for cuprate, iron-based, and low-T-c 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.