[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-119200-en":3,"doc-seo-119200-105":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},119200,8796095461610,"Oliver","https://ap-avatar.wpscdn.com/davatar_276721f389ce27ea32af1340a28f341c",8,"Research & Report","Practical Marketplace Optimization at Uber Using Causally-Informed Machine Learning - end-to-end ML and optimization for automated budget decisions","Marketplace budget allocation for ride-share platforms relies on managing incentives across drivers, riders, and regions, while meeting predefined budgets and maximizing business value under operational constraints. This work proposes an end-to-end machine learning and optimization pipeline that automates decision-making using causal deep learning, feature-store based training, and model serving. It combines S-Learner estimates, a tensor B-spline regression for efficiency shape control, and ADMM/primal-dual convex optimization with backtesting to improve resource allocation efficiency for Uber cities.","Practical Marketplace Optimization at Uber Using Causally-Informed Machine Learning  \narXiv :2407 . 19078v 1 [ cs .LG] 26 Jul 2024  \nBobby Chen∗† Siyu Chen∗ Jason Dowlatabadi∗ Yu Xuan Hong∗ Vinayak Iyer∗ Uday Mantripragada∗  \nRishabh Narang∗ Apoorv Pandey∗ Zijun Qin∗ Abrar Sheikh∗ Hongtao Sun∗† Jiaqi Sun∗ Uber Technologies, Inc San Francisco, CA, USA mhwalker, [zijunq@uber.com](zijunq@uber.com)  \nMatthew Walker∗ Kaichen Wei∗ Chen Xu∗ Jingnan Yang∗  \nAllen T. Zhang∗  \nGuoqing Zhang∗  \nABSTRACT  \nBudget allocation of marketplace levers, such as incentives for drivers to complete certain trips or promotions for riders to take more trips have long been both a technical and business challenge at Uber. It is crucial to understand the impact of lever budget changes on the market and to estimate their cost efficiency given the need to achieve predefined budgets, where the eventual goal is to find the optimal allocations under those constraints that maximize some objective of value to the business. In this paper, we introduce an end-to-end machine learning and optimization procedure to automate budget decision-making for cities where Uber operates. This procedure relies on a suite of applications, including feature store, model training and serving, optimizers and backtesting to measure the prediction and causal accuracy. We propose a state-of-the-art deep learning (DL) estimator based on S-Learner that leverages massive amount of user experimental and temporal-spatial observational data. We also built a novel tensor B-Spline regression model to enforce efficiency shape control while retaining the sophistication of the DL models’ response surface, and solved the high-dimensional optimization problem with Alternating Direction Method of Multipliers (ADMM) and primal-dual interior point convex optimization. This procedure has demonstrated substantial improvement in Uber’s ability to allocate resources efficiently.  \nCCS CONCEPTS  \n• Computing methodologies → Machine learning algorithms;  \n• General and reference → Measurement; • Mathematics of computing → Mathematical optimization.  \n∗ Authors listed alphabetically.†Formerly at Uber Technologies, Inc.  \nPermission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than the author(s) must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission [and/or a fee. Request permissions from permissions@acm.org](and/or a fee. Request permissions from permissions@acm.org).  \nKDD 2024 Workshop, August 25–26, 2024, Barcelona, Spain  \n© 2024 Copyright held by the owner/author(s) . Publication rights licensed to ACM. ACM ISBN 978-1-4503-XXXX-X/18/06. . . $15.00 [https://doi.org/XXXXXXX.XXXXXXX](https://doi.org/XXXXXXX.XXXXXXX)  \nKEYWORDS  \nCausal machine learning, deep neural network, convex optimization, shape-constrained, measurement  \nACM Reference Format:  \nBobby Chen, Siyu Chen[1], Jason Dowlatabadi[1], Yu Xuan Hong[1], Vinayak Iyer[1], Uday Mantripragada[1], Rishabh Narang[1], Apoorv Pandey[1], Zijun Qin[1], Abrar Sheikh[1], Hongtao Sun[1][2], Jiaqi Sun[1], Matthew Walker[1], Kaichen Wei[1], Chen Xu[1], Jingnan Yang[1], Allen T. Zhang[1], and Guoqing Zhang[1] . 2024. Practical Marketplace Optimization at Uber Using Causally-Informed Machine Learning. In Proceedings of 2nd Workshop on Causal Inference and Machine Learning in Practice (KDD 2024 Workshop) . ACM, New York, NY, USA, 10 pages. [https://doi.org/XXXXXXX.XXXXXXX](https://doi.org/XXXXXXX.XXXXXXX)  \n1 INTRODUCTION  \nEach year, Uber manages a multi-billion dollar marketplace. In the third quarter of 2023 alone [1], Uber’s mobility business recorded $17.9 billion in","cbCairKlScYm7Z6P","https://ap.wps.com/l/cbCairKlScYm7Z6P","pdf",1035128,1,10,"English","en",105,"# Abstract\n# Introduction\n## Automated marketplace budgeting and causal prediction\n## Optimization under operational constraints\n# Core ML-Optimization Pipeline\n## Feature store, training, and serving\n## Causal estimators and efficiency modeling\n## Optimization method and backtesting\n# Methods and Models\n## S-Learner deep estimator\n## Tensor B-spline regression with shape control\n## ADMM and primal-dual interior point optimization\n# Evaluation and Results\n## Demonstrated improvement in allocation efficiency","[{\"question\":\"What problem does the paper address at Uber?\",\"answer\":\"It addresses how to allocate budgets across marketplace levers (e.g., incentives) to influence rider and driver behavior while achieving predefined budgets and maximizing business value.\"},{\"question\":\"How does the proposed system estimate the impact of budget changes?\",\"answer\":\"It uses causally-informed deep learning (based on an S-Learner) trained on historical experimental and temporal-spatial observational data to predict supply, demand, and a custom marketplace objective.\"},{\"question\":\"How is optimal budget allocation computed?\",\"answer\":\"The system builds a smoothing/efficiency model and solves the resulting high-dimensional optimization problem using ADMM and primal-dual interior point convex optimization, then backtests prediction and causal accuracy.\"}]","Practical Marketplace Optimization at Uber Using Causally-Informed Machine Learning - end-to-end ML and optimization for automated budget decisions | PDF",1785723058,25,{"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},"practical-marketplace-optimization-at-uber-using-causally-informed-machine-learning-end-to-end-ml-and-optimization-for-automated-budget-decisions","",{"@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/practical-marketplace-optimization-at-uber-using-causally-informed-machine-learning-end-to-end-ml-and-optimization-for-automated-budget-decisions/119200/",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","2026-08-05","2026-08-03",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},"What problem does the paper address at Uber?","Question",{"text":76,"@type":77},"It addresses how to allocate budgets across marketplace levers (e.g., incentives) to influence rider and driver behavior while achieving predefined budgets and maximizing business value.","Answer",{"name":79,"@type":74,"acceptedAnswer":80},"How does the proposed system estimate the impact of budget changes?",{"text":81,"@type":77},"It uses causally-informed deep learning (based on an S-Learner) trained on historical experimental and temporal-spatial observational data to predict supply, demand, and a custom marketplace objective.",{"name":83,"@type":74,"acceptedAnswer":84},"How is optimal budget allocation computed?",{"text":85,"@type":77},"The system builds a smoothing/efficiency model and solves the resulting high-dimensional optimization problem using ADMM and primal-dual interior point convex optimization, then backtests prediction and causal accuracy.","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},[94,98,102,106,111,116,121,124,129,132,135],{"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":125,"doc_module":4,"doc_module_name":46,"category_name":126,"show_sort_weight":127,"slug":128},9,"Religion & Spirituality",20,"religion-spirituality",{"id":127,"doc_module":4,"doc_module_name":46,"category_name":130,"show_sort_weight":127,"slug":131},"World Cup","world-cup",{"id":21,"doc_module":4,"doc_module_name":46,"category_name":133,"show_sort_weight":21,"slug":134},"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":46,"category_name":137,"show_sort_weight":107,"slug":138},19,"General","general"]