SciPost Submission Page
Sivers extraction with Neural Network
by I. P. Fernando, N. Newton, D. Seay & D. Keller
This is not the latest submitted version.
This Submission thread is now published as
Submission summary
| Authors (as registered SciPost users): | Ishara Fernando |
| Submission information | |
|---|---|
| Preprint Link: | scipost_202107_00108v1 (pdf) |
| Date submitted: | 2021-07-31 22:29 |
| Submitted by: | Fernando, Ishara |
| Submitted to: | SciPost Physics Proceedings |
| Proceedings issue: | 28th Annual Workshop on Deep-Inelastic Scattering (DIS) and Related Subjects (DIS2021) |
| Ontological classification | |
|---|---|
| Academic field: | Physics |
| Specialties: |
|
| Approaches: | Theoretical, Experimental, Computational |
Abstract
Psuedo-data with simulated experimental errors can be generated to train an ensemble of Artificial Neural Networks (ANN) implemented on a regression to extract Transverse Momentum-dependent Distributions (TMDs). A preliminary analysis is presented on the reliability in extraction of the Sivers function imposed in the pseudo-data given the bounds on the experimental errors, data sparsity, and complexity of phase-space.
Current status:
Reports on this Submission
Anonymous Report 1 on 2022-3-1 (Invited Report)
- Cite as: Anonymous, Report on arXiv:scipost_202107_00108v1, delivered 2022-03-01, doi: 10.21468/SciPost.Report.4571
Report
Their discussion of the fit results to HERMES2009 and HERMES2020 data states they are consistent, but the table of values shows parameters that are substantially incompatible. I suspect they mean to show that their neural network approach matches the observables despite the very different parameters for these two datasets. This may be something understood to a community expert, but is misleading to outsiders. A brief discussion of the two datasets, in particular what differs between them, and how the NN is expected to handle these differences, would be very helpful. The Journal's acceptance criteria for these proceedings are otherwise met.
If a revised manuscript is submitted, please also correct typos where found, eg the first word of the abstract (psuedo --> pseudo) and second-to-last word of the conclusions (Mlders --> Mulders).