Machine-learned metrics for predicting the likelihood of success in materials discovery
| Autor: | Yoolhee Kim, Bryce Meredig, Edward Kim, Erin Antono, Julia Ling |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2020 |
| Předmět: |
FOS: Computer and information sciences
Optimal design Computer Science - Machine Learning Computer science FOS: Physical sciences Machine learning computer.software_genre Machine Learning (cs.LG) Set (abstract data type) lcsh:TA401-492 General Materials Science Fraction (mathematics) Selection (genetic algorithm) lcsh:Computer software Condensed Matter - Materials Science business.industry Frame (networking) Critical question Materials Science (cond-mat.mtrl-sci) Computational Physics (physics.comp-ph) Computer Science Applications lcsh:QA76.75-76.765 Mechanics of Materials Modeling and Simulation lcsh:Materials of engineering and construction. Mechanics of materials Artificial intelligence Haystack business Physics - Computational Physics Design space computer |
| Zdroj: | npj Computational Materials, Vol 6, Iss 1, Pp 1-9 (2020) |
| ISSN: | 2057-3960 |
| DOI: | 10.1038/s41524-020-00401-8 |
| Popis: | Materials discovery is often compared to the challenge of finding a needle in a haystack. While much work has focused on accurately predicting the properties of candidate materials with machine learning (ML), which amounts to evaluating whether a given candidate is a piece of straw or a needle, less attention has been paid to a critical question: Are we searching in the right haystack? We refer to the haystack as the design space for a particular materials discovery problem (i.e. the set of possible candidate materials to synthesize), and thus frame this question as one of design space selection. In this paper, we introduce two metrics, the Predicted Fraction of Improved Candidates (PFIC), and the Cumulative Maximum Likelihood of Improvement (CMLI), which we demonstrate can identify discovery-rich and discovery-poor design spaces, respectively. Using CMLI and PFIC together to identify optimal design spaces can significantly accelerate ML-driven materials discovery. 13 pages, 10 figures, 2 tables |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|