Band Gap Predictor

Supports formulas with up to 4 elements

an ensemble of 30 GNN models

Try GaN, or ZrO2, or HfC_0.93, or Ni_10Fe_72Cr_18

The credit goes to Hsuan-Yu Kao and John Grimm for developing this model.

Cite this model:

Si-Da Xue and Qi-Jun Hong, Materials Properties Prediction (MAPP): Empowering the Prediction of Material Properties Solely Based on Chemical Formulas, Materials, 2024. 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.

Metrics

R2 training score: 0.92

R2 testing score: 0.88

Root mean square error, training: 0.50 eV

Root mean square error, testing: 0.63 eV

What is new

Version 1: ensemble of 30 deep learning models.

About this model

This model is based on Graph Neural Network (GNN) and Residual Neural Network (ResNet).
This ensemble model is based on bootstrap aggregating (bagging).

Video Introduction