CALPHAD Excess Parameter L(A,B;0)

CALPHAD parameter L, Ensemble of 30 GNN models, must have 2 elements

Try FeNi, NbW, or ScTh

Cite this model:

Qi-Jun Hong, Deep learning for CALPHAD modeling: Universal parameter learning solely based on chemical formula, arXiv, 2023. Download.

Qi-Jun Hong, A melting temperature database and a neural network model for melting temperature prediction, arXiv, 2021. Download.

Qi-Jun Hong, Sergey V Ushakov, Axel van de Walle, and Alexandra Navrotsky, Melting temperature prediction using a graph neural network model: From ancient minerals to new materials, PNAS, 2022. Download.

This model is currently deployed on Microsoft Azure and the Research Computing facilities at ASU. Due to network limit, computation may take up to 10 seconds.


R2 training score: 0.96

R2 testing score: 0.85

Root mean square error, training: 18 kJ/mol

Root mean square error, testing: 35 kJ/mol

What is new

Version 1: GNN + ensemble of 30 deep learning models + multi-task learning

About this model

This model is based on Graph Neural Network (GNN) and Residual Neural Network (ResNet).
This ensemble model employs bootstrap aggregating (bagging).
This ensemble model utilizes multi-task learning.

Video Introduction