2023 Fiscal Year Final Research Report
Discovery of innovative functional materials using state-of-the-art machine learning
| Project/Area Number |
19H01132
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (A)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Review Section |
Medium-sized Section 61:Human informatics and related fields
|
|---|
| Research Institution | The Institute of Statistical Mathematics |
|---|
Principal Investigator |
Yoshida Ryo 統計数理研究所, 先端データサイエンス研究系, 教授 (70401263)
|
|---|
| Project Period (FY) |
2019-04-01 – 2024-03-31
|
| Keywords | マテリアルズインフォマティクス / 機械学習 / データベース / シミュレーション / 高分子材料 / 準結晶 / 転移学習 |
|---|
| Outline of Final Research Achievements |
We have established the foundation (material database, theory and methodology of machine learning) of materials informatics for various material systems such as polymeric materials, inorganic compounds, and quasiperiodic materials. In particular, to overcome the problem of insufficient data resources, which is the biggest obstacle in data-driven materials research, we have promoted the integration of machine learning and computer experiments such as molecular dynamics simulations, the development of Sim2Real transfer learning methods for integrated analysis of heterogeneous data from real-world and computer experiments, and the development of materials database. We have also applied these methodologies to discover new materials for various material systems (quasicrystals, highly thermally conductive amorphous polymers, polymer liquid crystals, etc.), thus demonstrating the concept of materials informatics.
|
| Free Research Field |
マテリアルズインフォマティクス
|
| Academic Significance and Societal Importance of the Research Achievements |
様々な材料系を対象にマテリアルズインフォマティクスの方法・実践・実証に関する研究を実施した．特にデータ駆動型材料研究では，データ資源の不足が問題視されている．この問題を乗り越えるために，計算機実験による大規模データベースを構築し，転移学習等の方法論で大量のシミュレーションデータと限られた実験データを統合的に解析するすることで，高精度な予測器を構築できることを実証した．さらに，開発した機械学習の手法を用いて，準結晶，高熱伝導非晶質高分子，高分子液晶等，様々な新物質創製を実現した．なお全ての研究において論文発表時にデータとソースコードを公開した．
|
