Details
| Original language | English |
|---|---|
| Article number | 033106 |
| Number of pages | 13 |
| Journal | Physical Review A |
| Volume | 107 |
| Issue number | 3 |
| Early online date | 14 Mar 2023 |
| Publication status | Published - Mar 2023 |
Abstract
We investigate the application of deep learning to the retrieval of the internuclear distance in the two-dimensional H2+ molecule from the momentum distribution of photoelectrons produced by strong-field ionization. We study the effect of the carrier-envelope phase on the prediction of the internuclear distance with a convolutional neural network and investigate the possibility of reconstruction of the internuclear distance from one-dimensional momentum distributions. We apply the transfer learning technique to make our convolutional neural network applicable to distributions obtained for parameters outside the ranges of the training data. The convolutional neural network is compared with alternative approaches to this problem, including the direct comparison of momentum distributions, support-vector machines, and decision trees. These alternative methods are found to possess very limited transferability. Finally, we use the occlusion-sensitivity technique to extract the features that allow a neural network to make its decisions.
ASJC Scopus subject areas
- Physics and Astronomy(all)
- Atomic and Molecular Physics, and Optics
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
In: Physical Review A, Vol. 107, No. 3, 033106, 03.2023.
Research output: Contribution to journal › Article › Research › peer review
}
TY - JOUR
T1 - Transfer learning, alternative approaches, and visualization of a convolutional neural network for retrieval of the internuclear distance in a molecule from photoelectron momentum distributions
AU - Shvetsov-Shilovski, N. I.
AU - Lein, M.
N1 - Funding Information: We are grateful to S. Brennecke, F. Oppermann, and S. Yu for stimulating discussions. This work was supported by the Deutsche Forschungsgemeinschaft (Grant No. SH 1145/1-2).
PY - 2023/3
Y1 - 2023/3
N2 - We investigate the application of deep learning to the retrieval of the internuclear distance in the two-dimensional H2+ molecule from the momentum distribution of photoelectrons produced by strong-field ionization. We study the effect of the carrier-envelope phase on the prediction of the internuclear distance with a convolutional neural network and investigate the possibility of reconstruction of the internuclear distance from one-dimensional momentum distributions. We apply the transfer learning technique to make our convolutional neural network applicable to distributions obtained for parameters outside the ranges of the training data. The convolutional neural network is compared with alternative approaches to this problem, including the direct comparison of momentum distributions, support-vector machines, and decision trees. These alternative methods are found to possess very limited transferability. Finally, we use the occlusion-sensitivity technique to extract the features that allow a neural network to make its decisions.
AB - We investigate the application of deep learning to the retrieval of the internuclear distance in the two-dimensional H2+ molecule from the momentum distribution of photoelectrons produced by strong-field ionization. We study the effect of the carrier-envelope phase on the prediction of the internuclear distance with a convolutional neural network and investigate the possibility of reconstruction of the internuclear distance from one-dimensional momentum distributions. We apply the transfer learning technique to make our convolutional neural network applicable to distributions obtained for parameters outside the ranges of the training data. The convolutional neural network is compared with alternative approaches to this problem, including the direct comparison of momentum distributions, support-vector machines, and decision trees. These alternative methods are found to possess very limited transferability. Finally, we use the occlusion-sensitivity technique to extract the features that allow a neural network to make its decisions.
UR - http://www.scopus.com/inward/record.url?scp=85151160462&partnerID=8YFLogxK
U2 - 10.48550/arXiv.2211.01210
DO - 10.48550/arXiv.2211.01210
M3 - Article
AN - SCOPUS:85151160462
VL - 107
JO - Physical Review A
JF - Physical Review A
SN - 2469-9926
IS - 3
M1 - 033106
ER -