[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-127591-en":3,"doc-seo-127591-105":30,"detail-sidebar-cat-0-en-105":92},{"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":20,"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},127591,549768064622,"Anda","https://ap-avatar.wpscdn.com/davatar_6f874abed73319feea01a86fa6f0fab8",8,"Research & Report","An Optimal Policy for Target Localization with Application to Electron Microscopy - Paper Summary","This paper addresses finding the location of a target using a limited sequence of adaptive observations. Each observation queries an interval and runs an imperfect classifier whose reliability depends on the chosen cost and time allocation. A Bayesian objective combines posterior entropy with question cost, and the work proves that the one-step lookahead (greedy) policy is Bayes-optimal for any time horizon. The policy is computationally practical and is applied to mitochondrial localization in electron microscopy, where experiments show substantial acquisition-time reductions with comparable image quality.","An Optimal Policy for Target Localization with Application to Electron Microscopy  \nRaphael Sznitman  \nEcole Polytechnique Federale de Lausanne, Switzerland  \nAurelien Lucchi  \nEcole Polytechnique Federale de Lausanne, Switzerland  \nPeter I. Frazier  \nCornell University, Ithaca, NY 14850, USA  \nBruno M. Jedynak  \nJohns Hopkins University, Baltimore, MD 21218, USA  \nPascal Fua  \nEcole Polytechnique Federale de Lausanne, Switzerland  \n[raphael.sznitman@epfl.ch](raphael.sznitman@epfl.ch)  \n[aurelien.lucchi@epfl.ch](aurelien.lucchi@epfl.ch)  \n[pf98@cornell.edu](pf98@cornell.edu)  \n[bruno.jedynak@jhu.edu](bruno.jedynak@jhu.edu)  \n[pascal.fua@epfl.ch](pascal.fua@epfl.ch)  \nAbstract  \nThis paper considers the task of 􀀌nding a target location by making a limited number of sequential observations. Each observation results from evaluating an imperfect classi􀀌er of a chosen cost and accuracy on an interval of chosen length and position. Within a Bayesian framework, we study the problem of minimizing an objective that combines the entropy of the posterior distribution with the cost of the questions asked. In this problem, we show that the one-step lookahead policy is Bayes-optimal for any arbitrary time horizon. Moreover, this one-step lookahead policy is easy to compute and implement. We then use this policy in the context of localizing mitochondria in electron microscope images, and experimentally show that signi􀀌 -cant speed ups in acquisition can be gained, while maintaining near equal image quality at target locations, when compared to current policies.  \nProceedings of the 30 th International Conference on Machine Learning, Atlanta, Georgia, USA, 2013 . JMLR: W&CP volume 28 . Copyright 2013 by the author(s) .  \n1. Introduction  \nWe consider the problem of quickly identifying the location and size of an object of interest by asking a series of questions about the region it occupies. This is key to speeding up the imaging of target objects, such as intracellular structures, when using scanning electron microscopes (SEM), which can focus resources on promising areas (Lucchi et al. , 2011; Veeraraghavan et al. , 2010) . In this application, asking a question involves acquiring image data over a portion of the capture area and feeding it to a classi􀀌er that returns an estimate of the presence of a target within the queried region. This data can be either acquired quickly, which degrades the reliability of the estimate, or slowly, which severely impacts overall acquisition time.  \nWhile both image acquisition and classi􀀌cation are subject to noise, requiring the use of statistical inference to estimate the target's location, the heart of this work lies in the Decision Problem of what questions to ask to localize the target e􀀎ciently. As in Active Learning (Dasgupta et al. , 2007; Castro et al. , 2005; Settles, 2009) and Sequential Experimental Design (DeGroot, 1970; Wetherill & Glazebrook, 1986; Berry & Fristedt, 1985; Srinivas et al. , 2010), the decisions can be made adaptively, which may increase e􀀎ciency but makes the optimal decision policy more di􀀎cult to 􀀌nd.  \nIn this paper, we formulate this problem in a Bayesian  \nframework and explicitly characterize the globally Bayes-optimal policy. Furthermore, we prove that, somewhat surprisingly, the greedy, or one-step lookahead, policy is in fact Bayes-optimal over arbitrary time horizons. The greedy policy is easy to compute, which is bene􀀌cial in many time-sensitive and computationally-limited applications. This optimality result is in contrast with other solutions to Bayesian active learning problems, where the greedy policy is suboptimal and can only serve as a heuristic (Brochu et al. , 2009; Zhang et al. , 2003) .  \nWe demonstrate the e􀀋ectiveness of this optimal greedy policy in the context of SEM imaging, where our results show that images of desired quality at target locations are acquired in half the time required by other state-of-the-art methods.  \nAkin to this work, (Jedynak","cbCain7CEdPr7dJ8","https://ap.wps.com/l/cbCain7CEdPr7dJ8","pdf",2529829,1,9,"English","en",105,"# Introduction\n## Bayesian sequential decision framework\n## Related literature\n# Related Literature\n## Active learning and sequential experimental design","[{\"question\":\"How does the paper model sequential target localization?\",\"answer\":\"It models localization as making a limited number of adaptive observations, where each query evaluates an imperfect classifier over a chosen interval.\"},{\"question\":\"What is the key theoretical result about the policy?\",\"answer\":\"Within a Bayesian framework, the paper shows the one-step lookahead (greedy) policy is Bayes-optimal for any arbitrary time horizon.\"},{\"question\":\"How is the policy applied to electron microscopy?\",\"answer\":\"The method is used to localize mitochondria in electron microscope images by adaptively selecting queries, and experiments demonstrate significant speedups while maintaining near-equal image quality.\"}]","An Optimal Policy for Target Localization with Application to Electron Microscopy - Paper Summary | PDF",1785940152,23,{"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":87,"head_meta":89,"extra_data":91,"updated_unix":28},"an-optimal-policy-for-target-localization-with-application-to-electron-microscopy-paper-summary","",{"@graph":36,"@context":86},[37,54,69],{"@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/an-optimal-policy-for-target-localization-with-application-to-electron-microscopy-paper-summary/127591/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":63,"encodingFormat":61,"isAccessibleForFree":64,"interactionStatistic":65},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-23","2026-08-05",true,{"@type":66,"interactionType":67,"userInteractionCount":20},"InteractionCounter",{"@type":68},"ViewAction",{"@type":70,"mainEntity":71},"FAQPage",[72,78,82],{"name":73,"@type":74,"acceptedAnswer":75},"How does the paper model sequential target localization?","Question",{"text":76,"@type":77},"It models localization as making a limited number of adaptive observations, where each query evaluates an imperfect classifier over a chosen interval.","Answer",{"name":79,"@type":74,"acceptedAnswer":80},"What is the key theoretical result about the policy?",{"text":81,"@type":77},"Within a Bayesian framework, the paper shows the one-step lookahead (greedy) policy is Bayes-optimal for any arbitrary time horizon.",{"name":83,"@type":74,"acceptedAnswer":84},"How is the policy applied to electron microscopy?",{"text":85,"@type":77},"The method is used to localize mitochondria in electron microscope images by adaptively selecting queries, and experiments demonstrate significant speedups while maintaining near-equal image quality.","https://schema.org",{"og:url":52,"og:type":88,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":90,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":93},[94,98,102,106,111,116,121,124,128,131,135],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":95,"show_sort_weight":96,"slug":97},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":99,"show_sort_weight":100,"slug":101},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":103,"show_sort_weight":104,"slug":105},"Exam",70,"exam",{"id":107,"doc_module":4,"doc_module_name":46,"category_name":108,"show_sort_weight":109,"slug":110},5,"Comic",60,"comic",{"id":112,"doc_module":4,"doc_module_name":46,"category_name":113,"show_sort_weight":114,"slug":115},6,"Technology",50,"technology",{"id":117,"doc_module":4,"doc_module_name":46,"category_name":118,"show_sort_weight":119,"slug":120},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":122,"slug":123},30,"research-report",{"id":21,"doc_module":4,"doc_module_name":46,"category_name":125,"show_sort_weight":126,"slug":127},"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":107,"slug":138},19,"General","general"]