[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-126850-en":3,"doc-seo-126850-105":30,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},126850,1099523882367,"Hazel","https://ap-avatar.wpscdn.com/davatar_9964176cb1d06d4a9deccf72a44ae3dc",8,"Research & Report","Methods of machine learning for the analysis of cosmic rays mass composition with the KASCADE experiment data - research paper","The study investigates reconstructing the mass composition of high-energy cosmic rays using extensive air-shower experimental data. Multiple machine learning approaches are developed to recover energy spectra for individual primary nuclei between 1 and 100 PeV, leveraging public KASCADE data and Monte-Carlo simulations from the KCDC platform. The work evaluates uncertainties, including those from the unfolding procedure, and demonstrates that overall reconstruction accuracy surpasses the performance reported in earlier KASCADE analyses.","arXiv :2311 .06893v1 [ astro-ph .HE] 12 Nov 2023  \nMethods of machine learning for the analysis of cosmic rays mass composition with the KASCADE experiment data  \nM. Yu. Kuznetsova N. A. Petrova,b,c I. A. Plokhikha,b,d V. V. Sotnikova  \na Institute for Nuclear Research of the Russian Academy of Sciences, 117312, Moscow, Russia b Novosibirsk State University, 630090, Novosibirsk, Russia  \nc Budker Institute of Nuclear Physics, SB RAS, 630090, Novosibirsk, Russia  \nd Institute of Thermophysics, SB RAS, 630090, Novosibirsk, Russia E-mail: [mkuzn@inr.ac.ru](mkuzn@inr.ac.ru)  \nAbstract: We study the problem of reconstruction of high-energy cosmic rays mass composition from the experimental data of extensive air showers. We develop several machine learning methods for the reconstruction of energy spectra of separate primary nuclei at energies 1-100 PeV, using the public data and Monte-Carlo simulations of the KASCADE experiment from the KCDC platform. We estimate the uncertainties of our methods, including the unfolding procedure, and show that the overall accuracy exceeds that of the method used in the original studies of the KASCADE experiment.  \n\n| Contents\u003Cbr>1 Introduction\u003Cbr>2 Experiment, data and Monte-Carlo\u003Cbr>3 Machine learning methods\u003Cbr>3.1 Random Forest\u003Cbr>3.2 Multi-layer perceptron\u003Cbr>3.3 Convolutional Neural Network\u003Cbr>3.4 EfficientNet\u003Cbr>4 Performance of the methods\u003Cbr>4.1 Confusion matrices\u003Cbr>4.2 Reconstruction of mass components spectra\u003Cbr>4.3 Mass components reconstruction in mixtures\u003Cbr>5 Tests and uncertainties\u003Cbr>5.1 Ablation study\u003Cbr>5.2 Uncertainty of missed detectors in data\u003Cbr>5.3 Zenith angle dependence\u003Cbr>5.4 Energy dependence\u003Cbr>5.5 Cross-hadronic models reconstruction\u003Cbr>5.6 Test with the unblind set of the real data\u003Cbr>6 Unfolding and results\u003Cbr>6.1 Unfolding procedure\u003Cbr>6.2 Estimation of the uncertainties\u003Cbr>6.3 Comparison with the standard KASCADE reconstruction\u003Cbr>7 Conclusions\u003Cbr>A Details of the neural networks architectures\u003Cbr>B Details and tests for Random Forest model | 2\u003Cbr>3\u003Cbr>6 6\u003Cbr>6 6\u003Cbr>7\u003Cbr>7 7\u003Cbr>8 9\u003Cbr>11\u003Cbr>11\u003Cbr>11\u003Cbr>13\u003Cbr>13\u003Cbr>15\u003Cbr>16\u003Cbr>17\u003Cbr>18\u003Cbr>19\u003Cbr>23\u003Cbr>25\u003Cbr>25\u003Cbr>28 |\n| --- | --- |\n\n1 Introduction  \nMass composition is one of the main problems in the physics of high-energy cosmic rays (CR) [1] . Despite decades of experimental studies the precise mass composition above the so-called knee of the CR spectrum (E ≳ 1015 eV) is not known. While the general trend of composition becoming heavier with energy is expected from theory [2, 3], the results of the various cosmic rays experiments are incompatible with each other [4–6] . The knowledge of the spectra of separate mass components in the 1015 \u003C E \u003C 1018 eV energy range is important for the understanding of their origin, in particular, the transition between the galactic and the extragalactic cosmic rays is expected somewhere in this range [1] .  \nThe analysis of CR properties in this energy range is complicated by several issues. Unlike CRs of lower energies, these particles can be detected only indirectly, via so-called extensive air showers (EAS) of secondary particles, that they produce in the Earth’s atmosphere. The longitudinal and lateral structure of these showers can be recorded by various experimental techniques: Cherenkov and fluorescence light detection [6], detection of charged particles shower on the Earth surface and muon part of the shower underground [4, 5] . Then the properties of the primary particle can be reconstructed using this data. It is relatively simple to reconstruct the direction of the primary CR from the geometry of the shower and the CR energy from both the geometry and lateral distribution function (LDF) of secondary particles. At the same time, the accurate reconstruction of the primary particle type is much more difficult since the showers produced by different nuclei are not that different. One can either look for a longitudinal development of the shower in the atmosphere deriving the a","cbCailI85R8PfiU2","https://ap.wps.com/l/cbCailI85R8PfiU2","pdf",1753787,1,33,"English","en",105,"# Introduction\n## Cosmic-ray mass composition and the knee problem\n## Indirect detection via extensive air showers (EAS)\n## Role of hadronic interaction models\n## Goal and approach: modern reanalysis using machine learning\n# Experiment, data and Monte-Carlo\n# Machine learning methods\n## Random Forest\n## Multi-layer perceptron\n## Convolutional Neural Network\n## EfficientNet\n# Performance of the methods\n## Confusion matrices\n## Reconstruction of mass component spectra\n## Mixture reconstructions\n# Tests and uncertainties\n## Ablation study\n## Missed-detector uncertainty\n## Zenith-angle and energy dependence\n## Cross-hadronic models and unblind set tests\n# Unfolding and results\n## Unfolding procedure\n## Uncertainty estimation\n## Comparison with standard KASCADE reconstruction\n# Conclusions\n# Details and tests for models","[{\"question\":\"What reconstruction problem is addressed in this work?\",\"answer\":\"The work focuses on reconstructing the mass composition of high-energy cosmic rays from extensive air-shower experimental data.\"},{\"question\":\"Which energy range and primary nuclei are considered?\",\"answer\":\"It targets energy spectra of separate primary nuclei in the range of 1–100 PeV.\"},{\"question\":\"How are uncertainties and reconstruction reliability evaluated?\",\"answer\":\"The study estimates uncertainties, including those arising from the unfolding procedure, and compares overall accuracy against the standard KASCADE reconstruction approach.\"}]","Methods of machine learning for the analysis of cosmic rays mass composition with the KASCADE experiment data - research paper | PDF",1785935222,83,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"methods-of-machine-learning-for-the-analysis-of-cosmic-rays-mass-composition-with-the-kascade-experiment-data-research-paper","",{"@graph":36,"@context":85},[37,54,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/methods-of-machine-learning-for-the-analysis-of-cosmic-rays-mass-composition-with-the-kascade-experiment-data-research-paper/126850/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":62,"encodingFormat":61,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-05",true,{"@type":65,"interactionType":66,"userInteractionCount":4},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What reconstruction problem is addressed in this work?","Question",{"text":75,"@type":76},"The work focuses on reconstructing the mass composition of high-energy cosmic rays from extensive air-shower experimental data.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"Which energy range and primary nuclei are considered?",{"text":80,"@type":76},"It targets energy spectra of separate primary nuclei in the range of 1–100 PeV.",{"name":82,"@type":73,"acceptedAnswer":83},"How are uncertainties and reconstruction reliability evaluated?",{"text":84,"@type":76},"The study estimates uncertainties, including those arising from the unfolding procedure, and compares overall accuracy against the standard KASCADE reconstruction approach.","https://schema.org",{"og:url":52,"og:type":87,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":89,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":46,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":46,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":46,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":46,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]