Skip to content

Latest commit

 

History

History
32 lines (19 loc) · 1.43 KB

README.md

File metadata and controls

32 lines (19 loc) · 1.43 KB

SegrDNN

Codes for the paper entitled "Surface segregation machine-learned with inexpensive numerical fingerprint for the design of alloy catalysts" [Mol. Catal. 2023, 541, 113096].

Schematics

Closed-loop hyper-parameter tuning of DNN model using Bayesian optimization (BO) and 10-fold CV:

figure1

The flowchart above was re-designed based on the figure in the original paper [Mol. Catal. 2023, 541, 113096].

Included Codes

Hyper-parameter tuning

  • using_keras_tuner.py: for hyper-parameter tuning of DNN model for surface segregation energy (Esegr) link
  • run.sh: a shell script to prevent the python code from stopping link

Principal Component Analysis (PCA)

  • PCA_plots.ipynb link

Analyses based on predicted Esegr values

  • Esegr_vs_CN_SHAP_screening.ipynb link

Application of This Code

This code has been utilized in the following published paper:

  1. D. Shin, G. Choi, C. Hong, and J. W. Han, Mol. Catal. 2023, 541, 113096 (https://doi.org/10.1016/j.mcat.2023.113096)