Download file (182 KB)
Download as zip (177 KB)

Collection

Citation

Description:

Motivated by the observation that screening known materials using trained property models often fails to identify candidates that outperform those in the training set, we introduce a periodic descriptor-based exploration strategy to search for new, uncharted compositions. To support structure-aware validation, we constructed classification models for space group and Pearson symbol using approximately 150,000 experimentally reported compounds from the AtomWork-Adv. (AWA) database, developed by NIMS. These models enable structure prediction from formula-only inputs, thereby facilitating downstream simulations and structure generation.

Rights:

• In Copyright
  

Keyword: Inverse design, Structure prediction, Periodic descriptor

Conference: MRM2025 (2025-12-08 - 2025-12-13)

Funding:

Manuscript type: Not a journal article

MDR DOI: https://doi.org/10.48505/nims.6034

First published URL:

Related item:

Other identifier(s):

Contact agent:

Updated at: 2025-12-19 09:52:03 +0900

Published on MDR: 2025-12-19 14:11:37 +0900