A deep convolutional neural network (CNN) is developed to study symmetry energy () effects by learning the mapping between the symmetry energy and the two-dimensional (transverse momentum and rapidity) distributions of protons and neutrons in heavy-ion collisions. Supervised training is performed with labeled data-set from the ultrarelativistic quantum molecular dynamics (UrQMD) model simulation. It is found that, by using proton spectra on event-by-event basis as input, the accuracy for classifying the soft and stiff is about 60% due to large event-by-event fluctuations, while by setting event-summed proton spectra as input, the classification accuracy increases to 98%. The accuracies for 5-label (5 different ) classification task are about 58% and 72% by using proton and neutron spectra, respectively. For the regression task, the mean absolute errors (MAE) which measure the average magnitude of the absolute differences between the predicted and actual L (the slope parameter of ) are about 20.4 and 14.8 MeV by using proton and neutron spectra, respectively. Fingerprints of the density-dependent nuclear symmetry energy on the transverse momentum and rapidity distributions of protons and neutrons can be identified by convolutional neural network algorithm.
{ "license": [ { "url": "http://creativecommons.org/licenses/by/3.0/", "license": "CC-BY-3.0" } ], "copyright": [ { "holder": "The Author(s)", "statement": "The Author(s)", "year": "2021" } ], "control_number": "64949", "_oai": { "updated": "2021-10-26T06:43:26Z", "id": "oai:repo.scoap3.org:64949", "sets": [ "PLB" ] }, "authors": [ { "surname": "Wang", "given_names": "Yongjia", "affiliations": [ { "country": "China", "value": "School of Science, Huzhou University, Huzhou, China" } ], "full_name": "Wang, Yongjia", "orcid": "0000-0003-2506-0010", "email": "wangyongjia@zjhu.edu.cn" }, { "affiliations": [ { "country": "China", "value": "School of Science, Huzhou University, Huzhou, China" }, { "country": "China", "value": "School of Science, Zhejiang University of Technology, Hangzhou, China" } ], "surname": "Li", "given_names": "Fupeng", "full_name": "Li, Fupeng" }, { "affiliations": [ { "country": "China", "value": "School of Science, Huzhou University, Huzhou, China" }, { "country": "China", "value": "Institute of Modern Physics, Chinese Academy of Sciences, Lanzhou, China" } ], "surname": "Li", "email": "liqf@zjhu.edu.cn", "full_name": "Li, Qingfeng", "given_names": "Qingfeng" }, { "affiliations": [ { "country": "China", "value": "HiSilicon Research Department, Huawei Technologies Co., Ltd., Shenzhen, China" } ], "surname": "L\u00fc", "given_names": "Hongliang", "full_name": "L\u00fc, Hongliang" }, { "orcid": "0000-0001-9859-1758", "affiliations": [ { "country": "Germany", "value": "Frankfurt Institute for Advanced Studies, Frankfurt am Main, Germany" } ], "surname": "Zhou", "given_names": "Kai", "full_name": "Zhou, Kai" } ], "_files": [ { "checksum": "md5:6e1c8dca46dcf3131cce29e6b4c2cf73", "filetype": "xml", "bucket": "01d455e2-3e37-41cb-9fd1-429e01e0cf37", "version_id": "7e9351dd-b07b-428d-9d96-e86e73fa03b1", "key": "10.1016/j.physletb.2021.136669.xml", "size": 110366 }, { "checksum": "md5:aaca7a3965ddb9eb6045f7310ebc8dc8", "filetype": "pdf", "bucket": "01d455e2-3e37-41cb-9fd1-429e01e0cf37", "version_id": "28c65eb9-f1d4-4dc4-a2bd-5542a0ca20b3", "key": "10.1016/j.physletb.2021.136669.pdf", "size": 678754 }, { "checksum": "md5:e3f798c82256b6d24b2712842e009d82", "filetype": "pdf/a", "bucket": "01d455e2-3e37-41cb-9fd1-429e01e0cf37", "version_id": "c09f408c-2a60-4860-b314-e1dd20aeb9ed", "key": "10.1016/j.physletb.2021.136669_a.pdf", "size": 1028747 } ], "record_creation_date": "2021-09-28T15:30:22.647829", "titles": [ { "source": "Elsevier", "title": "Finding signatures of the nuclear symmetry energy in heavy-ion collisions with deep learning" } ], "collections": [ { "primary": "Physics Letters B" } ], "dois": [ { "value": "10.1016/j.physletb.2021.136669" } ], "publication_info": [ { "journal_volume": "822 C", "journal_title": "Physics Letters B", "material": "article", "artid": "136669", "year": 2021 } ], "$schema": "http://repo.scoap3.org/schemas/hep.json", "abstracts": [ { "source": "Elsevier", "value": "A deep convolutional neural network (CNN) is developed to study symmetry energy (<math><msub><mrow><mi>E</mi></mrow><mrow><mi>sym</mi></mrow></msub><mo>(</mo><mi>\u03c1</mi><mo>)</mo></math>) effects by learning the mapping between the symmetry energy and the two-dimensional (transverse momentum and rapidity) distributions of protons and neutrons in heavy-ion collisions. Supervised training is performed with labeled data-set from the ultrarelativistic quantum molecular dynamics (UrQMD) model simulation. It is found that, by using proton spectra on event-by-event basis as input, the accuracy for classifying the soft and stiff <math><msub><mrow><mi>E</mi></mrow><mrow><mi>sym</mi></mrow></msub><mo>(</mo><mi>\u03c1</mi><mo>)</mo></math> is about 60% due to large event-by-event fluctuations, while by setting event-summed proton spectra as input, the classification accuracy increases to 98%. The accuracies for 5-label (5 different <math><msub><mrow><mi>E</mi></mrow><mrow><mi>sym</mi></mrow></msub><mo>(</mo><mi>\u03c1</mi><mo>)</mo></math>) classification task are about 58% and 72% by using proton and neutron spectra, respectively. For the regression task, the mean absolute errors (MAE) which measure the average magnitude of the absolute differences between the predicted and actual L (the slope parameter of <math><msub><mrow><mi>E</mi></mrow><mrow><mi>sym</mi></mrow></msub><mo>(</mo><mi>\u03c1</mi><mo>)</mo></math>) are about 20.4 and 14.8 MeV by using proton and neutron spectra, respectively. Fingerprints of the density-dependent nuclear symmetry energy on the transverse momentum and rapidity distributions of protons and neutrons can be identified by convolutional neural network algorithm." } ], "imprints": [ { "date": "2021-10-26", "publisher": "Elsevier" } ], "acquisition_source": { "date": "2021-10-26T06:32:58.924452", "source": "Elsevier", "method": "Elsevier", "submission_number": "25232ed0362611ecaa1b7aa32592193b" } }