[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-124757-en":3,"doc-seo-124757-105":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},124757,1649267921044,"Ava Thompson","https://us-avatar.wpscdn.com/avatar/1800007509477c92dfb?_k=1782875107921204101",8,"Research & Report","Interpretable machine learning for finding intermediate-mass black holes","Definitive evidence that globular clusters (GCs) host intermediate-mass black holes (IMBHs) remains elusive. Machine learning models trained on GC simulations can suggest IMBH hosts from observable features, but face black-box complexity and biased training due to imperfect simulation physics or initial conditions. Explainability and out-of-distribution generalization are addressed using anchors for an XGBoost classifier and CORELS for a certifiably interpretable model. Real-data candidates are ranked using prediction confidence, OOD flags, and kernel-density similarity measures.","arXiv :2310 . 18560v1 [ astro-ph .GA] 28 Oct 2023  \nDraft version October 31, 2023  \nTypeset using LATEX default style in AASTeX631  \nInterpretable machine learning for finding intermediate-mass black holes  \nMario Pasquato , 1, 2, 3, 4 Piero Trevisan ,5, 6 Abbas Askar ,7 Pablo Lemos , 1, 3, 4, 8 Gaia Carenini,9 Michela Mapelli , 10, 2, 11, 12 and Yashar Hezaveh4, 1, 8  \n1 Département de Physique, Université de Montréal, Montreal, Quebec H3T 1J4, Canada  \n2 Physics and Astronomy Department Galileo Galilei, University of Padova, Vicolo dell’Osservatorio 3, I–35122, Padova, Italy  \n3 Mila - Quebec Artificial Intelligence Institute, Montreal, Quebec, Canada  \n4 Ciela, Computation and Astrophysical Data Analysis Institute, Montreal, Quebec, Canada  \n5 Department of Physics, Università di Roma Sapienza, Piazzale Aldo Moro 2, 00185 Rome, Italy  \n6 INAF-Osservatorio Astronomico di Roma, via Frascati 33, 00040 Monte Porzio Catone, Italy  \n7 Nicolaus Copernicus Astronomical Center, Polish Academy of Sciences, ul. Bartycka 18, 00-716 Warsaw, Poland  \n8 Center for Computational Astrophysics, Flatiron Institute, NY, USA  \n9 Département d’Informatique, École normale supérieure - PSL Research University, Paris, France.  \n10 Institut für Theoretische Astrophysik, Zentrüm für Astronomie, Albert-Ueberle-Strasse 2, D-69120, Heidelberg, Germany  \n11 INFN-Padova, Via Marzolo 8, I–35131 Padova, Italy  \n12 INAF - Osservatorio Astronomico di Padova, Vicolo dell’Osservatorio 5, I-35122 Padova, Italy  \n(Received XXX; Revised YYY; Accepted October 31, 2023)  \nSubmitted to ApJ  \nABSTRACT  \nDefinitive evidence that globular clusters (GCs) host intermediate-mass black holes (IMBHs) is elusive.  \nMachine learning (ML) models trained on GC simulations can in principle predict IMBH host candidates based on observable features. This approach has two limitations: first, an accurate ML model is expected to be a black box due to complexity; second, despite our efforts to realistically simulate GCs, the simulation physics or initial conditions may fail to fully reflect reality. Therefore our training data may be biased, leading to a failure in generalization on observational data. Both the first issue  \n-explainability/interpretability- and the second-out of distribution generalization and fairness- are active areas of research in ML. Here we employ techniques from these fields to address them: we use the anchors method to explain an XGBoost classifier; we also independently train a natively interpretable model using Certifiably Optimal RulE ListS (CORELS) . The resulting model has a clear physical meaning, but loses some performance with respect to XGBoost. We evaluate potential candidates in real data based not only on classifier predictions but also on their similarity to the training data, measured by the likelihood of a kernel density estimation model. This measures the realism of our simulated data and mitigates the risk that our models may produce biased predictions by working in extrapolation. We apply our classifiers to real GCs, obtaining a predicted classification, a measure of the confidence of the prediction, an out-of-distribution flag, a local rule explaining the prediction of XGBoost and a global rule from CORELS.  \n1. INTRODUCTION  \nThe detection of quasars at high redshift (see, e.g. , Mortlock et al. 2011 ; Schindler et al. 2023 ; Maiolino et al. 2023) requires a mechanism for the rapid assembly of supermassive black holes. Intermediate-mass black holes (IMBHs) bridge the gap between stellar-mass remnants and supermassive black holes, potentially playing an important role as seeds for the latter (see, e.g. , Woods et al. 2019 and Volonteri et al. 2021 for two recent reviews) . Searches for IMBHs in the present-day Universe are being actively carried out, with focus on high-density environments such as the Galactic center (Oka et al. 2016 ; Ballone et al. 2018 ; Takekawa et al. 2019b,a, 2020 ; Kaneko et al. 2023 ; The GRAVITY Collaboration et al","cbCaikj13MhwQlio","https://ap.wps.com/l/cbCaikj13MhwQlio","pdf",10725688,1,21,"English","en",105,"# Abstract\n# Introduction\n## IMBH formation and search environments\n## Indirect detection via dynamical effects","[{\"question\":\"Why is confirming intermediate-mass black holes in globular clusters difficult?\",\"answer\":\"Definitive evidence is elusive, and searches rely on indirect dynamical signatures that are not uniquely determined.\"},{\"question\":\"What limitations affect ML models trained on globular cluster simulations?\",\"answer\":\"They tend to behave as black boxes and can generalize poorly because simulated training data may be biased or not represent real initial conditions and physics.\"},{\"question\":\"How does the approach improve interpretability and reduce the risk of biased predictions?\",\"answer\":\"It explains an XGBoost classifier using anchors, trains a certifiably interpretable CORELS model, and evaluates real candidates using both prediction outputs and kernel-density similarity to training data.\"}]","Interpretable machine learning for finding intermediate-mass black holes | PDF",1785894321,53,{"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},"interpretable-machine-learning-for-finding-intermediate-mass-black-holes","",{"@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/interpretable-machine-learning-for-finding-intermediate-mass-black-holes/124757/",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},"Why is confirming intermediate-mass black holes in globular clusters difficult?","Question",{"text":75,"@type":76},"Definitive evidence is elusive, and searches rely on indirect dynamical signatures that are not uniquely determined.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"What limitations affect ML models trained on globular cluster simulations?",{"text":80,"@type":76},"They tend to behave as black boxes and can generalize poorly because simulated training data may be biased or not represent real initial conditions and physics.",{"name":82,"@type":73,"acceptedAnswer":83},"How does the approach improve interpretability and reduce the risk of biased predictions?",{"text":84,"@type":76},"It explains an XGBoost classifier using anchors, trains a certifiably interpretable CORELS model, and evaluates real candidates using both prediction outputs and kernel-density similarity to training data.","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"]