[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-123557-en":3,"doc-seo-123557-105":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},123557,7971461741311,"Ophelia","https://ap-avatar.wpscdn.com/avatar/74000253aff267980c6?x-image-process=image/resize,m_fixed,w_180,h_180&k=1779345379180704826",8,"Research & Report","Dynamic Programming Boosting for Discriminative Macro-Action Discovery - Research paper","Dynamic Programming Boosting for Discriminative Macro-Action Discovery addresses automatic macro-action discovery in imitation learning by casting the task as change-point detection. Instead of relying on generative change-point modeling, the approach uses discriminative learning to avoid joint distribution modeling. A supervised algorithm, DPBoost, extends classical boosting by combining it with dynamic programming and alternately refining strong predictors and changepoint estimates. Experiments evaluate tasks with naturally occurring change-points and demonstrate application to macro-action discovery for complex image-based goal-planning with large feature sets.","View metadata, citation and similar [papers at ](papers at core.ac.uk)[core.ac.uk](papers at core.ac.uk) brought to you by CORE  \nprovided by Infoscience- École polytechnique fédérale de Lausanne  \nDynamic Programming Boosting for Discriminative Macro-Action Discovery  \nLeonidas Lefakis 1 ;2 Franc¸ois Fleuret 1 ;2  \nLEONIDAS . LEFAKIS @IDIAP. CH FRANCOIS . FLEURET @IDIAP. CH  \n1Idiap Research Institute, Martigny, Switzerland  \n2´Ecole Polytechnique Fdrale de Lausanne, Lausanne, Switzerland  \nAbstract  \nWe consider the problem of automatic macroaction discovery in imitation learning, which we cast as one of change-point detection. Unlike prior work in change-point detection, the present work leverages discriminative learning algorithms.  \nOur main contribution is a novel supervised learning algorithm which extends the classical Boosting framework by combining it with dynamic programming. The resulting process alternatively improves the performance of individual strong predictors and the estimated changepoints in the training sequence.  \nEmpirical evaluation is presented for the proposed method on tasks where change-points arise naturally as part of a classification problem. Finally we show the applicability of the algorithm to macro-action discovery in imitation learning and demonstrate it allows us to solve complex image-based goal-planning problems with thousands of features.  \n1. Introduction  \nThe supervised learning framework provides a large variety of powerful tools capable of addressing the complexity and variability of a number of growing applications. In this framework, methods are typically developed under an i.i.d. assumption concerning the application data. Here however we seek to leverage supervised machine learning methods in a slightly different setting. We consider the data to be generated sequentially via a mixture model comprising a latent variable, given which the i.i.d. assumption holds. The value of this latent variable however is not known during the training phase and must be inferred.  \nProceedings of the 31 st International Conference on Machine Learning, Beijing, China, 2014 . JMLR: W&CP volume 32 . Copyright 2014 by the author(s) .  \nThis problem of latent variable inference in sequentially generated data can be seen as one of change-point detection (Fearnhead & Liu, 2007) . Given the sequentially generated data (X; Y ) 2 Rd 􀀂 f1:::Cg change-point detection consists in finding positions n in the sequence which mark changes in the nature of the joint probability p(X; Y ) . Change-points can then be seen as an abrupt change in the source generating the data. Such a change can concern changes either some of the parameters of the underlying model, or even to the nature of the model itself. To give a concrete example, which we will address in section 4.2, in the case of speech segmentation, a change-point in a conversation can signal either the change of speaker or a change in the nature of the background noise.  \nDue to the nature of the problem, most previous work on change-point detection has focused on generative approaches. Such approaches attempt to model the joint distribution p(X; Y ) and to detect changes in these models over the sequences. In the present work however we are concerned with classification problems which lend themselves more naturally to discriminative approaches. Thus we propose to take a discriminative view of the problem and instead look directly at the conditional probability distributions p (YjX) and more specifically at the decision function argmaxY p (Y = yjX) . This allows us to leverage discriminative classifiers, while avoiding the complex, and ultimately unnecessary, problem of joint distribution modeling (Ng & Jordan, 2001) .  \nOur main contribution is a novel method that automatically assigns samples in a training set to a mixture ofclassifiers using the concept of macro-classes. Given a training set, our proposed algorithm DPBoost (Dynamic Programming Boosting) comprises an","cbCaisqsOiSfl6Y2","https://ap.wps.com/l/cbCaisqsOiSfl6Y2","pdf",1828497,1,9,"English","en",105,"# Introduction\n# Related Work","[{\"question\":\"What problem does DPBoost target in imitation learning?\",\"answer\":\"DPBoost targets automatic macro-action discovery in imitation learning by reformulating it as a change-point detection problem.\"},{\"question\":\"How does the method differ from prior change-point detection work?\",\"answer\":\"Unlike generative approaches that model the joint distribution, this work uses a discriminative learning view based on conditional decision functions.\"},{\"question\":\"What is the core idea behind DPBoost’s training procedure?\",\"answer\":\"DPBoost extends boosting with dynamic programming and alternates between improving individual strong predictors and estimating macro-class assignments for samples, under temporal regularity assumptions.\"}]","Dynamic Programming Boosting for Discriminative Macro-Action Discovery - Research paper | PDF",1785817303,23,{"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":86,"head_meta":88,"extra_data":90,"updated_unix":28},"dynamic-programming-boosting-for-discriminative-macro-action-discovery-research-paper","",{"@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/dynamic-programming-boosting-for-discriminative-macro-action-discovery-research-paper/123557/",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 problem does DPBoost target in imitation learning?","Question",{"text":75,"@type":76},"DPBoost targets automatic macro-action discovery in imitation learning by reformulating it as a change-point detection problem.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does the method differ from prior change-point detection work?",{"text":80,"@type":76},"Unlike generative approaches that model the joint distribution, this work uses a discriminative learning view based on conditional decision functions.",{"name":82,"@type":73,"acceptedAnswer":83},"What is the core idea behind DPBoost’s training procedure?",{"text":84,"@type":76},"DPBoost extends boosting with dynamic programming and alternates between improving individual strong predictors and estimating macro-class assignments for samples, under temporal regularity assumptions.","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},[93,97,101,105,110,115,120,123,127,130,134],{"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":21,"doc_module":4,"doc_module_name":46,"category_name":124,"show_sort_weight":125,"slug":126},"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":46,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":46,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":46,"category_name":136,"show_sort_weight":106,"slug":137},19,"General","general"]