Query-and-Learn Machine Learning framework to model the stability mechanism of REFe12 magnets
| Project/Area Number |
21K14396
|
|---|
| Research Category |
Grant-in-Aid for Early-Career Scientists
|
| Allocation Type | Multi-year Fund |
|---|
| Review Section |
Basic Section 26010:Metallic material properties-related
|
|---|
| Research Institution | Japan Advanced Institute of Science and Technology |
|---|
Principal Investigator |
NGUYEN DuongNguyen 北陸先端科学技術大学院大学, 先端科学技術研究科, 助教 (20879978)
|
|---|
| Project Period (FY) |
2021-04-01 – 2024-03-31
|
| Project Status |
Granted (Fiscal Year 2022)
|
| Budget Amount *help |
¥4,420,000 (Direct Cost: ¥3,400,000、Indirect Cost: ¥1,020,000)
Fiscal Year 2023: ¥1,040,000 (Direct Cost: ¥800,000、Indirect Cost: ¥240,000)
Fiscal Year 2022: ¥1,690,000 (Direct Cost: ¥1,300,000、Indirect Cost: ¥390,000)
Fiscal Year 2021: ¥1,690,000 (Direct Cost: ¥1,300,000、Indirect Cost: ¥390,000)
|
|---|
| Keywords | materials discovery / rare-earths magnets / evidence-based theory / active learning / SmFe12 / materials exploration / REFe12 magnets |
|---|
| Outline of Research at the Start |
This research aim to model stability mechanisms and monitor the discovery process of RE(Fe1-x-yAxBy)12 magnets using Machine Learning (ML) with RE as rare-earth; A and B as Ga, Co, Mo, Cu, Al, and Ti substituted elements. We build a query-and-learn method comprising Active learning and mechanism-based similarity measurement to learn stability mechanism from the discovery’s feedback. Three results are expected: (1) model stability mechanism by ML, (2) unveil meaningful structure-stability correlations, and (3) monitor the discovery process of REFe12-substituted structures.
|
| Outline of Annual Research Achievements |
The research project has been process with three main achievements:
(1) develop an active learning framework in investigating the structure-stability of SmFe12 family magnets (published in MRS Bulletin Impact). (2) develop an Evidence-based similarity measure for materials science data (published in J. Appl. Phys.) (3) develop a Machine Learning-aided Genetic algorithm to investigate the stability SmFe12 based magnets (published in J. Appl. Phys.)
From data contribution aspect, published articles in (1) and (3) provided systematic quantum calculation datasets for SmFe12 magnets and framework for higher efficiency data calculation. From theoretical contribution aspect, article in (2) investigated a new method bridging between Dempster theory of evidence to investigate materials science data.
|
| Current Status of Research Progress |
Current Status of Research Progress
1: Research has progressed more than it was originally planned.
Reason
The grant focuses on building a query-and-learn framework to accelerate materials discovery process using interpretable machine learning systems. We clarify the process via the following points:
1- regarding the proposed query-and-learn framework: We have built two calculation infrastructures integrated machine learning and well-known quantum-based calculation methods (VASP and USPEX), which is capable of iteratively calculating new data to investigate a certain materials family
2- regarding accelerating the materials discovery process: lowering calculation resources was performed via our proposed framework
3- regarding interpretable ability: all proposed frameworks introduced methods for researchers to be able to monitor and understand the materials discovery process from multiple viewpoints
|
| Strategy for Future Research Activity |
In the future, we continue focus on improving the proposed framework to (1) increasing the efficiency of calculation resource (2) providing higher quality data via extracting more meaningful structure-stability relationship
|
Report
(2 results)
Research Products
(12 results)
-
-
-
-
-
[Journal Article] Evidence-based recommender system for high-entropy alloys2021
Author(s)
Minh-Quyet Ha, Duong-Nguyen Nguyen, Viet-Cuong Nguyen, Takahiro Nagata, Toyohiro Chikyow, Hiori Kino, Takashi Miyake, Thierry Denoeux, Van-Nam Huynh, Hieu-Chi Dam
-
Journal Title
Nature Computational Science
Volume: 1
Issue: 7
Pages: 470-478
DOI
Related Report
Peer Reviewed / Open Access / Int'l Joint Research
-
-
-
-
-
-
-
