[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-122614-en":3,"doc-seo-122614-105":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},122614,1099513958607,"Jiven","https://ap-avatar.wpscdn.com/avatar/100002390cf8733938c?x-image-process=image/resize,m_fixed,w_180,h_180&k=1778829742770036399",8,"Research & Report","Toward a taxonomy of trust for probabilistic machine learning - Research review summary","Probabilistic machine learning increasingly shapes decisions across medicine, economics, politics, and other domains. A taxonomy is proposed to pinpoint where trust in a data-analysis pipeline can fail: mapping real-world goals to training-data goals, translating abstract goals to concrete mathematical problems, solving those problems via algorithms, and executing them through specific code implementations. Two case studies illustrate failure points, and a broad set of approaches is discussed for strengthening trust at each stage.","Broderick, Tamara, Gelman, Andrew, Meager, Rachael, Smith, Anna L. & Zheng, Tian (2023) Toward a taxonomy of trust for probabilistic machine learning. Science Advances, 9(7). [https://doi.org/10.1126/sciadv.abn3999](https://doi.org/10.1126/sciadv.abn3999)  \n[https://researchonline.lse.ac.uk/id/eprint/118302/](https://researchonline.lse.ac.uk/id/eprint/118302/)  \nVersion: Published Version  \nLicence: Creative Commons: Attribution 4 .0  \nLSE Research Online is the repository for research produced by the London School of Economics and Political Science. For more information, please refer to our Policies  \npage or contact [lseresearchonline@lse.ac.uk](lseresearchonline@lse.ac.uk)  \nSCIENCE ADVANCES | REVIEW  \nCOMPUTER SCIENCE  \nToward a taxonomy of trust for probabilistic machine learning  \nTamara Broderick1*, Andrew Gelman2,3, Rachael Meager4, Anna L. Smith5, Tian Zheng2  \nProbabilistic machine learning increasingly informs critical decisions in medicine, economics, politics, and beyond. To aid the development of trust in these decisions, we develop a taxonomy delineating where trust in an analysis can break down: (i) in the translation of real-world goals to goals on a particular set of training data,(ii) in the translation of abstract goals on the training data to a concrete mathematical problem,(iii) in the use of an algorithm to solve the stated mathematical problem, and (iv) in the use of a particular code implementation of the chosen algorithm. We detail how trust can fail at each step and illustrate our taxonomy with two case studies. Finally, we describe a wide variety of methods that can be used to increase trust at each step of our taxonomy. The use of our taxonomy highlights not only steps where existing research work on trust tends to concentrate and but also steps where building trust is particularly challenging.  \nCopyright © 2023 The Authors, some rights reserved; exclusive licensee American Association for the Advancement of Science. No claim to original U.S. Government Works. Distributed under a Creative Commons Attribution License 4.0 (CC BY) .  \n“Science, at its core, is a social phenomenon. It is a reflection of people, of our relationships, and of our institutions. When we provide inputs to the algorithm, when we program the device, when we design, test, and research, we are making human choices—choices that bring our social world to bear in a new and powerful way.” —Alondra Nelson, Deputy Director for Science and Society, White House Office of Science and Technology Policy, 2021.  \nINTRODUCTION  \nMachine learning (ML) in general, and probabilistic methods in particular, are increasingly used to make major decisions in science, the social sciences, and engineering, with the potential to profoundly affect individuals’ day-to-day lives. For instance, probabilistic methods have driven knowledge of the spread and effects of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) (1– 3), underlie election predictions at the Economist (4, 5), and can guide our understanding of the efficacy of microloans in alleviating poverty (6). Given the large and growing impact of probabilistic ML, it behooves us to make sure that its outputs are useful for its users’stated purposes.  \nThere are many potential points, though, where a data analysis pipeline may break down. This issue becomes especially pressing as statistics and ML workflows grow increasingly complex to face modern challenges. These challenges arise not only from the sheer size of the data but also from the inherent difficulty of the problems being studied. \"Big data\" are messy data: confounded data rather than random samples, observational data rather than experiments, and available data rather than direct measurements of underlying constructs of interest. To make relevant inferences from big data, we need to extrapolate from sample to population, from control to treatment group, and from measurements to  \n1Department of Electrical Engineering and Computer","cbCair0bvENcxqPm","https://ap.wps.com/l/cbCair0bvENcxqPm","pdf",773986,1,13,"English","en",105,"# Introduction\n## Points of failure in probabilistic machine learning pipelines\n## Existing literature on trust and related properties\n## Proposed taxonomy and case studies","[{\"question\":\"What does the paper’s taxonomy of trust cover in probabilistic machine learning?\",\"answer\":\"It identifies where trust can break down across four stages: translating real-world goals to training-data goals, mapping abstract goals to mathematical problems, using an algorithm to solve them, and relying on a specific code implementation.\"},{\"question\":\"Why are trust issues increasingly important for probabilistic ML?\",\"answer\":\"Because probabilistic ML affects high-stakes decisions and the analysis pipeline becomes more complex, with many modeling assumptions and decision points that increase potential failure modes.\"},{\"question\":\"How does the paper evaluate and illustrate the taxonomy?\",\"answer\":\"It uses two case studies to show how trust can fail at the different steps and discusses methods that can increase trust at each stage.\"}]","Toward a taxonomy of trust for probabilistic machine learning - Research review summary | PDF",1785811732,33,{"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},"toward-a-taxonomy-of-trust-for-probabilistic-machine-learning-research-review-summary","",{"@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/toward-a-taxonomy-of-trust-for-probabilistic-machine-learning-research-review-summary/122614/",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","2026-08-04",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 does the paper’s taxonomy of trust cover in probabilistic machine learning?","Question",{"text":75,"@type":76},"It identifies where trust can break down across four stages: translating real-world goals to training-data goals, mapping abstract goals to mathematical problems, using an algorithm to solve them, and relying on a specific code implementation.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"Why are trust issues increasingly important for probabilistic ML?",{"text":80,"@type":76},"Because probabilistic ML affects high-stakes decisions and the analysis pipeline becomes more complex, with many modeling assumptions and decision points that increase potential failure modes.",{"name":82,"@type":73,"acceptedAnswer":83},"How does the paper evaluate and illustrate the taxonomy?",{"text":84,"@type":76},"It uses two case studies to show how trust can fail at the different steps and discusses methods that can increase trust at each stage.","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"]