This repository is the reproduction for "Machine Learning the Voltage of Electrode Materials in Metal-Ion Batteries".
├── dataset
├── preprocess # feature engineering scripts
├── regression # train scripts
├── .flake8
├── .gitignore
├── README.md
└── requirements.txt
Python version : 3.6.5
$ git clone https://github.com/nissy-dev/machine-learning-the-voltage-of-electrode-materials.git
$ cd machine-learning-the-voltage-of-electrode-materials
$ python -v venv venv
$ source venv/bin/activate
$ pip install -r requirements.txt
$ cd preprocess
$ python create_features.py
YYYY_MM_DD
is the date you created features. Please set your date.
$ cd regression
$ python train.py --feat-path ../dataset/feature/repro_features_YYYY_MM_DD.csv --method svr --out-dir result/svr
$ python train.py --feat-path ../dataset/feature/repro_features_YYYY_MM_DD.csv --method krr --out-dir result/krr
$ python train_dnn.py --feat-path ../dataset/feature/repro_features_YYYY_MM_DD.csv --out-dir result/dnn
SVR | KRR | DNN | *SVR | *KRR | *DNN | |
---|---|---|---|---|---|---|
fold 1 | 0.41 | 0.45 | 0.50 | 0.51 | 0.54 | 0.42 |
fold 2 | 0.43 | 0.44 | 0.52 | 0.25 | 0.28 | 0.48 |
fold 3 | 0.45 | 0.45 | 0.47 | 0.26 | 0.27 | 0.42 |
fold 4 | 0.38 | 0.41 | 0.46 | 0.35 | 0.47 | 0.44 |
fold 5 | 0.42 | 0.42 | 0.46 | 0.38 | 0.43 | 0.44 |
fold 6 | 0.40 | 0.41 | 0.46 | 0.62 | 0.71 | 0.42 |
fold 7 | 0.40 | 0.41 | 0.44 | 0.43 | 0.42 | 0.43 |
fold 8 | 0.42 | 0.42 | 0.46 | 0.59 | 0.62 | 0.42 |
fold 9 | 0.40 | 0.46 | 0.47 | 0.53 | 0.57 | 0.45 |
fold 10 | 0.41 | 0.42 | 0.51 | 0.28 | 0.30 | 0.48 |
10 fold MAE (mean±std) | 0.42±0.02 | 0.43±0.02 | 0.48±0.03 | 0.42±0.13 | 0.46±0.14 | 0.43±0.03 |
H-test | 0.45 | 0.47 | 0.54 | 0.40 | 0.39 | 0.43 |
Na-test | (0.63) | (0.59) | (0.61) | 1.00 | 0.93 | 1.25 |
*SVR, *KRR, *DNN are reported in the paper.
[1] Rajendra P. JoshiJesse, et al. Machine Learning the Voltage of Electrode Materials in Metal-Ion Batteries. ACS Appl. Mater. Interfaces, 18494-18503, 2019.
[2] Anubhav Jain, Shyue Ping Ong, Geoffroy Hautier, Wei Chen, William Davidson Richards, Stephen Dacek, Shreyas Cholia, Dan Gunter, David Skinner, Gerbrand Ceder, et al. Commentary: The materials project: A materials genome approach to accelerating materials innovation. Apl Materials, 1(1):011002, 2013