2025 · Computational Materials Science

QuantumShellNet: Ground-state eigenvalue prediction of materials using electronic shell structures and fermionic properties via convolutions

Can Polat , Hasan Kurban , Mustafa Kurban

Research highlights

QuantumShellNet: learning from electronic shells

Predict ground-state eigenvalues from electronic shell representations and fermionic properties.

Why we investigated it

Fast property models need representations that retain relevant electronic information.

What the work contributes

QuantumShellNet combines an orbital encoder with a physics-informed neural network.

What the evidence shows

The paper reports comparisons with DFT and neural-wavefunction methods across its evaluated elements and molecules. Those comparisons apply to the reported benchmark regime.

Build on this work

Explore the QuantumShellNet repository and paper to study the orbital representation and reproduce the documented prediction workflow.

  • QuantumShellNet
  • Ground-state eigenvalues
  • Electronic shells
  • Quantum machine learning
Cite
@article{polat2025quantumshellnet,
  title = {QuantumShellNet: Ground-state eigenvalue prediction of materials using electronic shell structures and fermionic properties via convolutions},
  author = {Can Polat and Hasan Kurban and Mustafa Kurban},
  year = 2025,
  journal = {Computational Materials Science},
  doi = {10.1016/j.commatsci.2024.113366},
}