Saved in:
| Main Author: | |
|---|---|
| Format: | Recurso digital |
| Language: | |
| Published: |
Zenodo
2026
|
| Online Access: | https://doi.org/10.5281/zenodo.18733451 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Table of Contents:
- <p>This repository presents a Hybrid Tabular–Graph Neural Network framework for predicting battery intercalation voltages. The model integrates chemically informed tabular descriptors, including Mendeleev and Matminer features, with crystal structure–derived graph representations processed through a Transformer-based Graph Neural Network architecture. By combining redox-aware compositional features with structure-driven attention mechanisms, the framework captures both chemical and spatial dependencies governing voltage behavior. The workflow includes data preprocessing, feature engineering, structure retrieval, model training, benchmarking against baseline models, and reproducible evaluation scripts.</p>