Target Material Property‐Dependent Cluster Analysis of Inorganic Compounds

The cluster analysis of materials categorizes them according to similarities based on the features of materials, providing insight into the relationship between the materials. Conventional cluster analyses typically use basic features derived from the chemical composition and crystal structure without considering target properties such as band gap and dielectric constant. However, such approaches do not meet demands for grading materials according to properties of interest simultaneously with chemical and structural similarities. In this article, we propose a clustering method grouping similar materials in terms of both the target properties and features. We compare the clustering considering the cohesive energy with that considering the band gap of metal oxides, showing that their categorizations are clearly different. We further analyze several clusters classified by the band gap and reveal coordination environments related to each range of the band gap. The clustering for the electronic static dielectric constant identifies a cluster involving several perovskite-type oxides and balancing with the band gap near the Pareto front. Our method enables analyses with different viewpoints from those of the conventional clustering and feature importance analyses by taking the relationship between the target property and the features into account.