[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-122072-en":3,"doc-seo-122072-105":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},122072,8796095461564,"Liam","https://ap-avatar.wpscdn.com/davatar_155a257f0dc6eb9ab79c44ca47cae57d",8,"Research & Report","Client-Aided Privacy-Preserving Machine Learning - Protocols and Techniques","Privacy-preserving machine learning (PPML) enables multiple distrustful parties to jointly train machine learning models on private data while revealing no information beyond the final trained models. This work studies a client-aided two-server setting where two non-colluding servers train collaboratively using data held by many clients. Efficient protocols are developed for linear regression, logistic regression, and neural networks, including secure inner product, sign checks, activation functions, and division on secret-shared values, with security strengthened from semi-honest to malicious.","Client-Aided Privacy-Preserving Machine Learning  \nPeihan Miao 1 , Xinyi Shi 1 , Chao Wu2 , and Ruofan Xu3  \n1 Brown University, Providence, USA  \n2 University of California, Riverside, USA  \n3 University of Illinois Urbana-Champaign, Urbana, USA  \nAbstract  \nPrivacy-preserving machine learning (PPML) enables multiple distrusting parties to jointly train ML models on their private data without revealing any information beyond the final trained models. In this work, we study the client-aided two-server setting where two non-colluding servers jointly train an ML model on the data held by a large number of clients. By involving the clients in the training process, we develop efficient protocols for training algorithms including linear regression, logistic regression, and neural networks. In particular, we introduce novel approaches to securely computing inner product, sign check, activation functions (e.g. , ReLU, logistic function), and division on secret shared values, leveraging lightweight computation on the client side. We present constructions that are secure against semi-honest clients and further enhance them to achieve security against malicious clients. We believe these new client-aided techniques may be of independent interest.  \nWe implement our protocols and compare them with the two-server PPML protocols presented in SecureML (Mohassel and Zhang, S&P’17) across various settings and ABY2.0 (Patra et al., Usenix Security’21) theoretically. We demonstrate that with the assistance of untrusted clients in the training process, we can significantly improve both the communication and computational efficiency by orders of magnitude. Our protocols compare favorably in all the training algorithms on both LAN and WAN networks.  \nKeywords: Privacy-Preserving Machine Learning, Secure Multi-Party Computation, ClientAided Protocols.  \n1 Introduction  \nIn recent years, we have witnessed machine learning (ML) emerge as one of the most influential technologies and rapidly expanding research domains. Its applications span a diverse spectrum, ranging from recommendation systems to self-driving cars, large language models, and even medical prediction and diagnosis. This is in part due to increasing amount of data being collected and available in the Big Data era. Meanwhile, as these machine learning algorithms and applications are deployed in various real-world scenarios, data privacy is becoming increasingly critical, especially in domains dealing with sensitive or confidential data such as healthcare, finance, and government. In cases where entities are hesitant or restricted from sharing their data due to privacy regulations, the significance of protecting data privacy is further emphasized.  \nAddressing these concerns, privacy-preserving machine learning (PPML) has become a crucial approach to training ML models in a distributed manner, which enables multiple distrusting parties  \nto collaboratively train ML models on their private data while maintaining data privacy. The most commonly considered setting in PPML, as proposed by Mohassel and Zhang [29], involves data owners (e.g., clients) secret sharing their data among two non-colluding parties (e.g., servers), who then jointly perform training on the secret-shared data.  \nAt a high level, this approach can be conceptualized as two servers engaging in secure two-party computation to train the ML model on secret-shared data. Importantly, the servers learn nothing beyond the final trained model, ensuring the privacy of individual data points. Nevertheless, prior work [16, 21 , 27 , 29 , 30 , 32 , 33] has overlooked the fact that the data was initially owned by the clients in the clear. In this work, we show that actively involving clients in the training process can yield significant improvements in both communication and computational efficiency of the overall protocol.  \n1.1 Our Contributions  \nWe study two-server PPML training where the data is held by a large number of clients. S","cbCaiey0B02MW68Z","https://ap.wps.com/l/cbCaiey0B02MW68Z","pdf",549514,1,42,"English","en",105,"# Abstract\n# Introduction\n## Our Contributions\n## Experimental Results","[{\"question\":\"What problem does client-aided PPML address in this paper?\",\"answer\":\"It enables multiple distrustful parties to train ML models on private client data without revealing anything beyond the final trained model, while improving efficiency by involving clients in computations.\"},{\"question\":\"Which ML training algorithms are covered by the proposed client-aided protocols?\",\"answer\":\"The paper presents protocols for linear regression, logistic regression, and neural networks.\"},{\"question\":\"What secure computation primitives are improved using lightweight client-side help?\",\"answer\":\"It introduces client-aided approaches for securely computing inner products, sign checks, activation functions (e.g., ReLU and logistic), and division on secret-shared values.\"}]","Client-Aided Privacy-Preserving Machine Learning - Protocols and Techniques | PDF",1785808691,106,{"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},"client-aided-privacy-preserving-machine-learning-protocols-and-techniques","",{"@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/client-aided-privacy-preserving-machine-learning-protocols-and-techniques/122072/",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","2026-08-04",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 client-aided PPML address in this paper?","Question",{"text":76,"@type":77},"It enables multiple distrustful parties to train ML models on private client data without revealing anything beyond the final trained model, while improving efficiency by involving clients in computations.","Answer",{"name":79,"@type":74,"acceptedAnswer":80},"Which ML training algorithms are covered by the proposed client-aided protocols?",{"text":81,"@type":77},"The paper presents protocols for linear regression, logistic regression, and neural networks.",{"name":83,"@type":74,"acceptedAnswer":84},"What secure computation primitives are improved using lightweight client-side help?",{"text":85,"@type":77},"It introduces client-aided approaches for securely computing inner products, sign checks, activation functions (e.g., ReLU and logistic), and division on secret-shared values.","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},[94,98,102,106,111,116,121,124,129,132,136],{"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":133,"doc_module":4,"doc_module_name":46,"category_name":134,"show_sort_weight":133,"slug":135},10,"Lifestyle","lifestyle",{"id":137,"doc_module":4,"doc_module_name":46,"category_name":138,"show_sort_weight":107,"slug":139},19,"General","general"]