Exploring Materials Band Structure Space with unsupervised Machine Learning.

NUÑEZ, MATIAS

COMPUTATIONAL MATERIALS SCIENCE

ELSEVIER SCIENCE BV

Lugar: Amsterdam; Año: 2019 vol. 158 p. 117 - 123

0927-0256

An unsupervised machine learning algorithm is applied for the first time to explore the space ofmaterials electronic band structures . T-student stochastic neighbor embedding (t-SNE), a state ofthe art algorithm for visualization of high dimensional data, is applied on feature spaces constructedby extracting electronic fingerprints straight from Brillouin zone of the materials. Different spaces aredesigned and mapped to lower dimensions allowing to analyze and explore this previously unchartedband structure space for thousands of materials at once. In all cases analyzed machine learning wasable to learn and cluster the materials depending on the features involved. t-SNE promises to be aextremely useful tool for exploring the materials space.