[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-119875-en":3,"doc-seo-119875-105":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},119875,1649267921044,"Ava Thompson","https://us-avatar.wpscdn.com/avatar/1800007509477c92dfb?_k=1782875107921204101",8,"Research & Report","Multi-Epoch Machine Learning for Galaxy Formation","This thesis applies machine learning to hydrodynamical cosmological simulations to address key challenges in galaxy formation modeling, including scale tradeoffs and disentangling coupled physical processes. A novel tree-based model predicts baryonic properties of dark-matter-only subhalos across wide redshift ranges, outperforming prior approaches and enabling analysis via feature-importance scores. Using LEGACY N-body data, it builds quasar mock catalogs and tests black-hole growth against observations, showing limitations in IllustrisTNG subgrid physics. Further chapters compare subgrid physics across simulations and parameter suites, and finish with neural-network plus symbolic-regression semi-analytic modeling for accurate galaxy populations.","This thesis has been submitted in fulfilment of the requirements for a postgraduate degree (e. g. PhD, MPhil, DClinPsychol) at the University of Edinburgh. Please note the following terms and conditions of use:  \n• This work is protected by copyright and other intellectual property rights, which are retained by the thesis author, unless otherwise stated.  \n• A copy can be downloaded for personal non-commercial research or study, without prior permission or charge.  \n• This thesis cannot be reproduced or quoted extensively from without first obtaining permission in writing from the author.  \n• The content must not be changed in any way or sold commercially in any format or medium without the formal permission of the author.  \n• When referring to this work, full bibliographic details including the author, title, awarding institution and date of the thesis must be given.  \nMulti-Epoch Machine Learning for Galaxy Formation  \nRobert J. McGibbon  \nO  \nF  \nI N B  \nU  \nE  \nD  \nR  \nG  \nH  \nDoctor of Philosophy The University of Edinburgh July 2023  \n2  \nLay summary  \nA multitude of complex physical processes are involved in galaxy formation and evolution. In recent years computers have become powerful enough to simulate representative volumes of the Universe and are able to reproduce a number of properties of observed galaxies. However, many challenges still remain in the field. One significant issue is the range of sizes which need to be considered, as processes which are important in shaping galaxies vary from individual stars up to the scale of the Universe itself. Thus, a tradeoff arises between accurately modeling small-scale phenomena and simulating large cosmic volumes. Another obstacle lies in the complex interplay between the different processes involved, which can make it difficult to distinguish the specific factors responsible for determining a particular galaxy property. In this thesis I demonstrate how machine learning can help to alleviate some of these problems. Machine learning is a field that enables computers to discern patterns directly from data, bypassing the need for explicit human instruction. By applying machine learning techniques to the data generated by galaxy simulations, I aim to address the tensions mentioned above.  \nIn the first part of my thesis I introduce a model that can be used to produce galaxy catalogs which span huge volumes of the universe. This method runs many times faster than a standard simulation. I show how my method is an improvement on previous work, and then use the catalog it generates to compare with observations of quasars at early times in the Universe.  \nThe subsequent chapters explore the ability of machine learning to provide insights into a simulation. I present two novel methods to do this, and use them both to compare different simulations. In one instance I focus on unraveling the mechanisms driving the buildup of stellar mass in galaxies. In the second case I investigate the flow of gas into and out of galaxies, exploring its influence on the growth of black holes.  \nii  \nAbstract  \nIn this thesis I utilise a range of machine learning techniques in conjunction with hydrodynamical cosmological simulations. In Chapter 2 I present a novel machine learning method for predicting the baryonic properties of dark matter only subhalos taken from N-body simulations. The model is built using a treebased algorithm and incorporates subhalo properties over a wide range of redshiftsas its input features. I train the model using a hydrodynamical simulation which enables it to predict black hole mass, gas mass, magnitudes, star formation rate, stellar mass, and metallicity. This new model surpasses the performance of previous models. Furthermore, I explore the predictive power of each input property by looking at feature importance scores from the tree-based model. By applying the method to the LEGACY N-body simulation I generate a large volume mock catalog of the quasar population at 􀁉 = 3","cbCaia1Y2vlBxZAo","https://ap.wps.com/l/cbCaia1Y2vlBxZAo","pdf",16599007,1,193,"English","en",105,"# Lay summary\n## Fast galaxy catalog generation and quasar comparison\n## Insights from machine learning on simulation processes\n# Abstract\n## Chapter 2: Predicting baryonic properties of subhalos\n## Chapter 3: Comparing galaxy evolution across simulations\n## Final chapter: Neural networks and symbolic regression semi-analytic model","[{\"question\":\"What is the main goal of the thesis?\",\"answer\":\"To demonstrate how machine learning can mitigate challenges in galaxy formation studies by learning patterns from simulation data and providing insights into simulated processes.\"},{\"question\":\"How does the thesis predict galaxy-related properties from dark-matter-only simulations?\",\"answer\":\"It introduces a novel tree-based machine learning model that uses subhalo properties across a wide redshift range to predict baryonic quantities such as black hole mass, gas mass, magnitudes, star formation rate, stellar mass, and metallicity.\"},{\"question\":\"What comparison does the thesis perform to test black-hole growth models?\",\"answer\":\"It generates a large-volume mock quasar catalog from the LEGACY N-body simulation and compares it with observations, concluding that the IllustrisTNG subgrid model does not accurately capture the growth of the most massive objects.\"}]","Multi-Epoch Machine Learning for Galaxy Formation | PDF",1785726766,486,{"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},"multi-epoch-machine-learning-for-galaxy-formation","",{"@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/multi-epoch-machine-learning-for-galaxy-formation/119875/",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","2026-08-03",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 is the main goal of the thesis?","Question",{"text":75,"@type":76},"To demonstrate how machine learning can mitigate challenges in galaxy formation studies by learning patterns from simulation data and providing insights into simulated processes.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does the thesis predict galaxy-related properties from dark-matter-only simulations?",{"text":80,"@type":76},"It introduces a novel tree-based machine learning model that uses subhalo properties across a wide redshift range to predict baryonic quantities such as black hole mass, gas mass, magnitudes, star formation rate, stellar mass, and metallicity.",{"name":82,"@type":73,"acceptedAnswer":83},"What comparison does the thesis perform to test black-hole growth models?",{"text":84,"@type":76},"It generates a large-volume mock quasar catalog from the LEGACY N-body simulation and compares it with observations, concluding that the IllustrisTNG subgrid model does not accurately capture the growth of the most massive objects.","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"]