Abstract
One of the most challenging problems in condensed matter physics is to predict crystal structure just from the chemical formula of the material. In this work, we present a robust machine learning (ML) predictor for the crystal point group of ternary materials (A_lB_mC_n) - as first step to predict the structure - with very small set of ionic and positional fundamental features. From ML perspective, the problem is strenuous due to multi-labelity, multi-class, and data imbalance. The resulted prediction is very reliable as high balanced accuracies are obtained by different ML methods. Many similarity-based approaches resulted in a balanced accuracy above 95% indicating that the physics is well captured by the reduced set of features; namely, stoichiometry, ionic radii, ionization energies, and oxidation states for each of the three elements in the ternary compound. The accuracy is not limited by the approach; but rather by the limited data points and we should expect higher accuracy prediction by having more reliable data.
Highlights
One of the most challenging problems in condensed matter physics is to predict crystal structure just from the chemical formula of the material
The data are extracted by matching each chemical formula in NOMAD repository with the generated material space of ternary materials, which was generated using most of the elements along with their possible oxidation states
We present a robust machine learning method to predict the crystal point group of ternary materials (Al BmCn ) and with very small set of needed ionic and positional fundamental features
Summary
One of the most challenging problems in condensed matter physics is to predict crystal structure just from the chemical formula of the material. In the past few years, materials science utilizing the vast accessible data through what is known as “material informatics” has witnessed a considerable growth[1,2,3,4,5,6,7,8,9,10] when compared to other related scientific activities such as theory, experiment, and computation. This could provide novel and unusual means to breakthroughs in accelerated materials discovery and competitive technological developments. In three-dimensional (3D) space, there are 32 point groups compared to 230 space g roups[22]
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have