[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-123056-en":3,"doc-seo-123056-105":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},123056,8796095461610,"Oliver","https://ap-avatar.wpscdn.com/davatar_276721f389ce27ea32af1340a28f341c",8,"Research & Report","Cross-Temporal Forecast Reconciliation at Digital Platforms with Machine Learning","Platform businesses require accurate high-dimensional forecast streams across multiple cross-sectional and temporal aggregation levels, along with coherence of forecasts across the full hierarchy to align decisions on pricing, product, controlling, and strategy. Because platform data streams involve complex characteristics and interdependencies, the paper proposes a non-linear hierarchical forecast reconciliation method using popular machine learning models. The approach produces cross-temporal reconciled forecasts in an automated and direct way and is fast enough for high-frequency, forecast-driven platform decisions. Empirical evaluation uses large-scale streaming datasets from a European on-demand delivery platform and a New York City bicycle-sharing system.","arXiv :2402 .09033v2 [ econ .EM] 31 May 2024  \nCross-Temporal Forecast Reconciliation at Digital Platforms with Machine Learning ∗  \nJeroen Romboutsa , Marie Ternesb , Ines Wilmsb  \naEssec Business School, France  \nb Maastricht University, School of Business and Economics, The Netherlands  \nJune 3, 2024  \nAbstract. Platform businesses operate on a digital core and their decision making requires high-dimensional accurate forecast streams at different levels of cross-sectional (e.g., geographical regions) and temporal aggregation (e.g., minutes to days) . It also necessitates coherent forecasts across all levels of the hierarchy to ensure aligned decision making across different planning units such as pricing, product, controlling and strategy. Given that platform data streams feature complex characteristics and interdependencies, we introduce anon-linear hierarchical forecast reconciliation method that produces cross-temporal reconciled forecasts in a direct and automated way through the use of popular machine learning methods. The method is sufficiently fast to allow forecast-based high-frequency decision making that platforms require. We empirically test our framework on unique, large-scale streaming datasets from a leading on-demand delivery platform in Europe and a bicycle sharing system in New York City.  \nKeywords: Hierarchical time series, Forecast reconciliation, Machine learning, Cross-temporal aggregation, Demand forecasting, Platform econometrics  \n∗ Correspondence to Ines Wilms, Maastricht University, School of Business and Economics, P.O. Box 616, 6200 MD Maastricht, The Netherlands, Email: [i.wilms@maastrichtuniversity.nl. We](i.wilms@maastrichtuniversity.nl. We) are very grateful to Benjamin Wolter, Pablo Perez Piskunow, Roger Caminal and Afonso Rodrigues for expert advice, to Rob Hyndman, Artem Prokhorov and Roberto Ren`o for comments provided on earlier versions of the paper, and to the participants of the IMS International Conference on Statistics and Data Science (ICSDS 2023) for helpful discussions. The last author was financially supported by the Dutch Research Council (NWO) under grant number VI.Vidi.211.032 .  \n1 Introduction  \nTime series to be forecasted are oftentimes naturally part of a hierarchical structure, where higher frequency and granular series are added together to form lower frequency aggregated series. Separate forecasts of each series rarely conserve this hierarchy, and forecast reconciliation methods are therefore required. In this paper, we consider novel forecast reconciliation methods for on-demand delivery platforms that require forecasts at different levels of cross-sectional and temporal aggregation. We introduce non-linear forecast reconciliation based on popular machine learning methods to capture the complex interdependencies of platform data.  \nPlatforms such as Uber, Lyft, GrubHub, UberEats or DoorDash are nowadays omnipresent in the global economy. They operate in high frequency and on a large number of verticals (regions, product categories, etc.) . Indeed, their market place is typically split up in different geographical regions, so a natural cross-sectional aggregation scheme arises from many individual delivery areas over zones towards few market places. Moreover, a temporal aggregation scheme naturally arises since, on the granular end, fast operational decisions (think in terms of minutes) are needed to ensure the platform’s service couriers are at the right time and location to serve consumer demand promptly and to determine compensation schemes for couriers through dynamic pricing. On the coarser end, strategic business decisions also require long-term planning since the budget available for each delivery area is set typically using daily demand forecasts. Accurate and coherent, i.e. reconciled, demand forecasts across all levels of the cross-sectional and temporal hierarchy are therefore key to the business’ success and to support aligned decision making across diffe","cbCaiaOqp4y9Ax7o","https://ap.wps.com/l/cbCaiaOqp4y9Ax7o","pdf",3995899,1,64,"English","en",105,"# Introduction\n## Cross-sectional and temporal aggregation in platform forecasting\n## Forecast reconciliation literature overview\n## Methodological foundations and related theoretical results","[{\"question\":\"Why do platform businesses need cross-temporal forecast reconciliation?\",\"answer\":\"They must produce forecasts that are accurate at different cross-sectional and temporal aggregation levels while remaining coherent across the hierarchy, so decisions in pricing, product, controlling, and strategy stay aligned.\"},{\"question\":\"What problem does the paper address in existing reconciliation approaches?\",\"answer\":\"Traditional methods often focus on cross-sectional coherence, while cross-temporal frameworks are less developed; the paper targets reconciliation when forecasts must be coherent across both temporal and cross-sectional hierarchies for platform data.\"},{\"question\":\"How does the proposed method achieve reconciled forecasts?\",\"answer\":\"It introduces a non-linear hierarchical forecast reconciliation method that directly and automatically generates cross-temporal reconciled forecasts by leveraging popular machine learning techniques.\"}]","Cross-Temporal Forecast Reconciliation at Digital Platforms with Machine Learning | PDF",1785814432,161,{"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},"cross-temporal-forecast-reconciliation-at-digital-platforms-with-machine-learning","",{"@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/cross-temporal-forecast-reconciliation-at-digital-platforms-with-machine-learning/123056/",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},"Why do platform businesses need cross-temporal forecast reconciliation?","Question",{"text":75,"@type":76},"They must produce forecasts that are accurate at different cross-sectional and temporal aggregation levels while remaining coherent across the hierarchy, so decisions in pricing, product, controlling, and strategy stay aligned.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"What problem does the paper address in existing reconciliation approaches?",{"text":80,"@type":76},"Traditional methods often focus on cross-sectional coherence, while cross-temporal frameworks are less developed; the paper targets reconciliation when forecasts must be coherent across both temporal and cross-sectional hierarchies for platform data.",{"name":82,"@type":73,"acceptedAnswer":83},"How does the proposed method achieve reconciled forecasts?",{"text":84,"@type":76},"It introduces a non-linear hierarchical forecast reconciliation method that directly and automatically generates cross-temporal reconciled forecasts by leveraging popular machine learning techniques.","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"]