{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/from-dark-matter-to-galaxies-with","title":"From Dark Matter to Galaxies with Convolutional Networks","arxiv_id":"1902.05965","date":"2019-02-15","proceeding":null,"authors":["Xinyue Zhang","Yanfang Wang","Wei zhang","Yueqiu Sun","Siyu He","Gabriella Contardo","Francisco Villaescusa-Navarro","Shirley Ho"],"abstract":"Cosmological surveys aim at answering fundamental questions about our\nUniverse, including the nature of dark matter or the reason of unexpected\naccelerated expansion of the Universe. In order to answer these questions, two\nimportant ingredients are needed: 1) data from observations and 2) a\ntheoretical model that allows fast comparison between observation and theory.\nMost of the cosmological surveys observe galaxies, which are very difficult to\nmodel theoretically due to the complicated physics involved in their formation\nand evolution; modeling realistic galaxies over cosmological volumes requires\nrunning computationally expensive hydrodynamic simulations that can cost\nmillions of CPU hours. In this paper, we propose to use deep learning to\nestablish a mapping between the 3D galaxy distribution in hydrodynamic\nsimulations and its underlying dark matter distribution. One of the major\nchallenges in this pursuit is the very high sparsity in the predicted galaxy\ndistribution. To this end, we develop a two-phase convolutional neural network\narchitecture to generate fast galaxy catalogues, and compare our results\nagainst a standard cosmological technique. We find that our proposed approach\neither outperforms or is competitive with traditional cosmological techniques.\nCompared to the common methods used in cosmology, our approach also provides a\nnice trade-off between time-consumption (comparable to fastest benchmark in the\nliterature) and the quality and accuracy of the predicted simulation. In\ncombination with current and upcoming data from cosmological observations, our\nmethod has the potential to answer fundamental questions about our Universe\nwith the highest accuracy.","url_abs":"http://arxiv.org/abs/1902.05965v2","url_pdf":"http://arxiv.org/pdf/1902.05965v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"from-dark-matter-to-galaxies-with","repo_url":"https://github.com/xz2139/From-Dark-Matter-to-Galaxies-with-Convolutional-Networks","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":null,"task_name":"CPU"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1902.05965","atlas_url":"https://app.syntology.ai/?focus=1902.05965","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}