search

anhender / mse_ML_datasets

MIT License
15 stars 2 forks source link

readme

Developing Benchmark Data for Efficient Comparison of Machine Learning Models for Materials Discovery

Related Data in Brief paper: "Benchmark datasets incorporating diverse tasks, sample sizes, material systems, and data heterogeneity for materials informatics"

Data

This repository contains datasets gathered from 16 previously published papers related to machine learning models used for materials discovery (termed Literature Data), as well as from the 2018 Materials Project database (termed MP Data).

Literature Data

The raw_data directory contains the pre-split datasets, while the split_data directory contains the split datasets. All datasets are in .csv format. An explanation of the specific method used for splitting each dataset is described below.

Datasets with more than 100 values were split into train, test, and validation sets, either using 10-fold or 5-fold cross-valdiation with the use of the scikit-learn K-Fold cross-validation method. The number of folds (10 vs 5) was determined primarily by following what each respective study did. Datasets with less than 100 values were split into train and test sets using the scikit-learn Leave-One-Out cross-validation method.

Each dataset can be downloaded and read in by using Pandas. An example is shown below.

import pandas as pd
data = pd.read_csv("Bala_classification_dataset/val/val0.csv")

The code that was used to split the datasets via the K-Fold and the Leave-One-Out cross-validation methods is located in the code directory. To run this code, the raw_data directory will need to be downloaded, and the path given at the beginning of the code will need to be changed.

def read_in_data():
   # the following path needs to be changed
    path = '/Users/anhender/Documents/UROP/Literature_Data/Lit_cleaned_data/'

MP Data

The Materials Project datasets are located in the Materials_Project_data directory. Seven material properties are available: bulk modulus, formation energy, energy above the convex hull, elastic anisotropy, band gap, magnetization, and shear modulus. All datasets are in .csv format. The data has been split into train-val-test splits with a 70-15-15 split.

These datasets can be downloaded in a similar manner as with the Literature Data, via Pandas (see above).

Works Cited

Below are the papers whose data were used to populate this repository, as well as a reference for the Materials Project. The corresponding datasets are listed below their respective paper. Additionally, a general overview of the papers can be seen in the Lit_Data_Overview_ver_final.xlsx file, which includes information such as the materials system, material property tested, machine learning model used, and performance metrics.

  1. Balachandran, Prasanna V., et al. "Experimental search for high-temperature ferroelectric perovskites guided by two-step machine learning." Nature communications 9.1 (2018): 1-9. https://doi.org/10.1038/s41467-018-03821-9
    • Bala_classification_dataset
    • Bala_regression_dataset
  2. Carrete, Jesús, et al. "Finding unprecedentedly low-thermal-conductivity half-Heusler semiconductors via high-throughput materials modeling." Physical Review X 4.1 (2014): 011019. https://doi.org/10.1103/PhysRevX.4.011019
    • Carrete_therm_conduct_train_clean
  3. Lee, Joohwi, et al. "Prediction model of band gap for inorganic compounds by combination of density functional theory calculations and machine learning techniques." Physical Review B 93.11 (2016): 115104. https://doi.org/10.1103/PhysRevB.93.115104
    • Lee_band_gaps
  4. Li, Wei, Ryan Jacobs, and Dane Morgan. "Predicting the thermodynamic stability of perovskite oxides using machine learning models." Computational Materials Science 150 (2018): 454-463. https://doi.org/10.1016/j.commatsci.2018.04.033
    • Li_DFT_and_features_clean
    • Li_DFT_dataset_clean
  5. Liu, Yue, et al. "The onset temperature (Tg) of AsxSe1-x glasses transition prediction: A comparison of topological and regression analysis methods." Computational Materials Science 140 (2017): 315-321. https://doi.org/10.1016/j.commatsci.2017.09.008
    • Liu_Tg_AsSe_glass
  6. Mannodi-Kanakkithodi, Arun, et al. "Machine learning strategy for accelerated design of polymer dielectrics." Scientific reports 6 (2016): 20952. https://doi.org/10.1038/srep20952
    • Mannodi_polymer_dielec
  7. Pilania, Ghanshyam, et al. "Accelerating materials property predictions using machine learning." Scientific reports 3.1 (2013): 1-6. https://doi.org/10.1038/srep02810
    • Pilania_Polymers_data
    • Pilania_Polymers_data_Spring_Const_clean
    • Pilania_Polymers_data_total_Dielec_Const_clean
  8. Pilania, Ghanshyam, et al. "Machine learning bandgaps of double perovskites." Scientific reports 6 (2016): 19375. https://doi.org/10.1038/srep19375
    • Pilania_double_perovskites_clean
  9. Pilania, Ghanshyam, and X-Y. Liu. "Machine learning properties of binary wurtzite superlattices." Journal of materials science 53.9 (2018): 6652-6664. https://doi.org/10.1007/s10853-018-1987-z
    • Pilania_superlattices
    • Pilania_superlattices_Formation_E_clean
    • Pilania_superlattices_GGA_Band_Gap_clean
    • Pilania_superlattices_HSE_Band_Gap_clean
    • Pilania_superlattices_elastic_consts
  10. Rajan, Arunkumar Chitteth, et al. "Machine-learning-assisted accurate band gap predictions of functionalized MXene." Chemistry of Materials 30.12 (2018): 4031-4038. https://doi.org/10.1021/acs.chemmater.8b00686
    • Rajan_MXene_data
  11. Seko, Atsuto, et al. "Machine learning with systematic density-functional theory calculations: Application to melting temperatures of single-and binary-component solids." Physical Review B 89.5 (2014): 054303. https://doi.org/10.1103/PhysRevB.89.054303
    • Seko_melt_temps
  12. Wei, Han, et al. "Predicting the effective thermal conductivities of composite materials and porous media by machine learning methods." International Journal of Heat and Mass Transfer 127 (2018): 908-916. https://doi.org/10.1016/j.ijheatmasstransfer.2018.08.082
    • Wei_composite_materials
    • Wei_porous_media
  13. Wu, K., et al. "Prediction of polymer properties using infinite chain descriptors (ICD) and machine learning: Toward optimized dielectric polymeric materials." Journal of Polymer Science Part B: Polymer Physics 54.20 (2016): 2082-2091. https://doi.org/10.1002/polb.24117
    • Wu_DFT_Eg_dielec_consts
    • Wu_Exp_Tg
    • Wu_Exp_dielec_const
    • Wu_loss_tang_100Hz
    • Wu_loss_tang_1kHz
  14. Xue, Dezhen, et al. "Accelerated search for materials with targeted properties by adaptive design." Nature communications 7.1 (2016): 1-9. https://doi.org/10.1038/ncomms11241
    • Xue_thermal_hysteresis
  15. Zeng, Shuming, et al. "Machine learning-aided design of materials with target elastic properties." The Journal of Physical Chemistry C 123.8 (2019): 5042-5047. https://doi.org/10.1021/acs.jpcc.9b01045
    • Zeng_elastic_prop
  16. Zhuo, Ya, Aria Mansouri Tehrani, and Jakoah Brgoch. "Predicting the band gaps of inorganic solids by machine learning." The journal of physical chemistry letters 9.7 (2018): 1668-1673. https://doi.org/10.1021/acs.jpclett.8b00124
    • Zhuo_classification_data
    • Zhuo_regression_data
  17. A. Jain, S.P. Ong, G. Hautier, W. Chen, W.D. Richards, S. Dacek, S. Cholia, D. Gunter, D. Skinner, G. Ceder, K.A. Persson (* = equal contributions). The Materials Project: A materials genome approach to accelerating materials innovation APL Materials, 2013, 1(1), 011002. doi:10.1063/1.4812323

Acknowledgments

This work was supported by the University of Utah's Undergraduate Reserach Opportunities Program (UROP) for the Fall 2020 Semester. We also gratefully acknowledge support from the National Science Foundation CAREER Award, grant number 1651668.