Neural network approaches for quantum energy level prediction

Alireza Khalili Golmankhaneh; Roman Pasechnik; Palle E. T. Jørgensen

doi:10.1007/s00500-026-11300-3

Back

Neural network approaches for quantum energy level prediction

Journal article

Peer reviewed

Neural network approaches for quantum energy level prediction

Alireza Khalili Golmankhaneh, Roman Pasechnik and Palle E. T. Jørgensen

Soft computing (Berlin, Germany)

05/07/2026

DOI: 10.1007/s00500-026-11300-3

View Online

Abstract

Artificial neural networks (ANNs) are employed to predict the energy levels of quantum systems, including the hydrogen atom, the harmonic oscillator, and the particle in a box, using various activation functions. The results demonstrate that the choice of activation function significantly influences prediction accuracy. This data-driven framework provides a powerful alternative to explicitly solving the Schrödinger equation, enabling direct estimation of quantum energy spectra from experimental data. The proposed method offers an efficient and flexible tool for the analysis of quantum systems in laboratory settings.

Details

Title: Subtitle: Neural network approaches for quantum energy level prediction
Creators: Alireza Khalili Golmankhaneh
Roman Pasechnik - Lund University
Palle E. T. Jørgensen - University of Iowa
Resource Type: Journal article
Publication Details: Soft computing (Berlin, Germany)
DOI: 10.1007/s00500-026-11300-3
ISSN: 1432-7643
eISSN: 1433-7479
Publisher: Springer Nature
Language: English
Electronic publication date: 05/07/2026
Academic Unit: Mathematics
Record Identifier: 9985164081202771

Metrics

1 Record Views