| About this Abstract |
| Meeting |
2024 TMS Annual Meeting & Exhibition
|
| Symposium
|
Computational Discovery and Design of Materials
|
| Presentation Title |
Materials Discovery via Machine Learning on Li-based Battery Materials |
| Author(s) |
Suchismita Goswami |
| On-Site Speaker (Planned) |
Suchismita Goswami |
| Abstract Scope |
Considerable efforts have been made to discover novel materials using machine learning techniques for identifying similar potential material with known properties. Li Based Compounds are good candidates as solid-state electrolytes for solid-state lithium-ion batteries due to high conductivities. We use chemical and structural descriptors generated from the ICSD database on thousands of Li-based compounds to find potential materials around a user defined compound using distance measure in both featurized and reduced dimensions. For the identification of novel materials, neighborhood maps of materials are usually generated employing the dimensionality reduction algorithms. However, a significant information loss after the dimensionality reduction in 2-dimensions as compared to the high dimensional featurized space has been observed. Here we implement a different approach that will reduce the observed difference. In this presentation, we will discuss machine learning methods, and results on neighborhood maps on Li-based battery materials and compounds. |
| Proceedings Inclusion? |
Planned: |