aimat-lab / ml4heos

Machine Learning methods for high entropy oxides

ML4HEOs

Machine Learning methods for high entropy oxides

Paper

Phase-property diagrams for multicomponent oxide systems towards material libraries
Leonardo Velasco, Juan S. Castillo, Kante M. Veerraju, Pascal Friederich, Horst Hahn
Universidad Nacional de Colombia
Technical University Darmstadt
Karlsruhe Institute of Technology

Contact

• High entropy alloys, experiments: [email protected]
• Machine learning: [email protected]

Requirements

• python 3.x
• Tensorflow 2 and Keras

Data

• All data used from our paper can be found in the data directory

Machine Learning

• All python scripts can be found in the code directory
• The convolutional neural networks described in the paper can be trained using the train_NN_xrd.py script
• New synthetic data using reference XRD diffractograms can be generated using the generate_synthetic_data.py script