[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-117221-en":3,"doc-seo-117221-105":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},117221,5909877438554,"Maeve","https://ap-avatar.wpscdn.com/avatar/5600025385ad2bf12a7?_k=1778553567797529272",8,"Research & Report","Advances in Differential Privacy and Differentially Private Machine Learning","Differential privacy (DP) research and privacy-preserving machine learning have expanded rapidly, covering new variants, accounting techniques, and practical implementations in data analysis systems used by organizations such as census bureaus. Prior surveys often emphasize applications within specific contexts like data publishing, individual learning tasks, unstructured data analysis, and location privacy. This survey instead focuses on recent theoretical developments, including Renyi DP and Concentrated DP, new mechanisms and techniques, and detailed advances in differentially private machine learning, along with real-world privacy-preserving deployments.","arXiv :2404 .04706v 1 [ cs .CR] 6 Apr 2024  \nAdvances in Differential Privacy and Differentially Private Machine Learning  \nSaswat Das 1[0000−0002−6126−1699] and Subhankar Mishra 1[0000−0002−9910−7291]  \nNational Institute of Science Education and Research, an OCC of Homi Bhabha National Institute, India {saswat.das,[smishra](smishra}@niser.ac.in)[}](smishra}@niser.ac.in)[@niser.ac.in](smishra}@niser.ac.in)  \nAbstract. There has been an explosion of research on differential privacy (DP) and its various applications in recent years, ranging from novel variants and accounting techniques in differential privacy to the thriving  \nfield of differentially private machine learning (DPML) to newer implementationsin practice, like those by various companies and organisations such as census bureaus. Most recent surveys focus on the applications of differential privacy in particular contexts like data publishing, specific machine learning tasks, analysis of unstructured data, location privacy etc. This work  \nthus seeks to fill the gap for a survey that primarily discusses recent developments in the theory of differential privacy along with newer DP variants, viz. Renyi DP and Concentrated DP, novel mechanisms and techniques, and the theoretical developments in differentially private machine learning in proper detail. In addition, this survey discusses its applications to privacy-preserving machine learning in practice and a few practical implementations of DP.  \nKeywords: Differential Privacy · Privacy-Preserving Machine Learning  \n· Trustworthy AI  \n1 Introduction  \nThe explosion of popularity and adoption of fields like machine learning and big data, and powerful data processing machinery has meant that high quality data is considered to be among the most valuable, high utility commodities. This data, which often includes sensitive details about certain individuals and entities, helps demographers draw useful information about a population and socioeconomic distribution across an area of land, helps tech companies analyse the usage habits of and issues faced by users to design updates to their products, and helps medical professionals to improve upon diagnostic systems and medical care, to understand diseases better, create medical data visualisations, etc. Companies like Netflix and YouTube often utilise data to provide personalised content recommendations for their users.  \nBut as a corollary, this has enabled the extraction of certain, potentially sensitive, information about the individuals in databases (a.k.a. data subjects) unless  \n2 Das and Mishra  \nprotected in some form. This sensitive information can be used to the detriment of the concerned data subjects by entities like insurance companies that could use sensitive data on whether someone has a particular ailment or habit to increase their insurance premiums or deny them insurance and thus violate legislations like HIPAA that deal with such concerns about sensitive medical data, by other individuals or agencies to blackmail them or track their activities/movements, by governments or political agencies to gain sensitive data on citizens etc. This has naturally led to privacy concerns. Well known attacks like the linkage attack on the medical records released by the Massachusetts Group Insurance Commission[107], and that that on the Netflix Prize database[82] respectively compromised the medical records of government employees in the state of Massachusetts in the 1990s and the private content consumption data of Netflix viewers in 2006 . Very prominently the reconstruction attack on the 2010 US Census data[45] was able to reconstruct the private microdata of a significant proportion of American citizens from deidentified and publicly available census data. Kasiviswanathan, Rudelson, and Smith[64](2012) demonstrated that linear reconstruction attacks can be successful in various, including some seemingly “non-linear”, settings, including when applied to a large class of ERM a","cbCaiks01T6dCA7m","https://ap.wps.com/l/cbCaiks01T6dCA7m","pdf",854660,1,45,"English","en",105,"# Introduction\n## Privacy risks and attacks\n## Statistical disclosure limitation (SDL) and anonymity notions\n## From vulnerability to formal privacy guarantees","[{\"question\":\"What problem does the document address in the context of data and privacy?\",\"answer\":\"It addresses how sensitive information about individuals can be inferred from high-utility datasets and how this leads to privacy concerns, including well-known reconstruction and linkage attacks.\"},{\"question\":\"What privacy-preserving techniques are discussed as earlier approaches?\",\"answer\":\"The document references statistical disclosure limitation (SDL) methods and anonymity concepts such as k-anonymity and its variants like t-closeness, l-diversity, and m-invariance.\"},{\"question\":\"Which differential privacy theories and variants does the survey emphasize?\",\"answer\":\"It emphasizes recent theoretical developments including Renyi DP and Concentrated DP, along with new mechanisms and techniques and advances in differentially private machine learning.\"}]","Advances in Differential Privacy and Differentially Private Machine Learning | PDF",1785674467,113,{"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},"advances-in-differential-privacy-and-differentially-private-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/advances-in-differential-privacy-and-differentially-private-machine-learning/117221/",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","2026-08-02",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 the document address in the context of data and privacy?","Question",{"text":75,"@type":76},"It addresses how sensitive information about individuals can be inferred from high-utility datasets and how this leads to privacy concerns, including well-known reconstruction and linkage attacks.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"What privacy-preserving techniques are discussed as earlier approaches?",{"text":80,"@type":76},"The document references statistical disclosure limitation (SDL) methods and anonymity concepts such as k-anonymity and its variants like t-closeness, l-diversity, and m-invariance.",{"name":82,"@type":73,"acceptedAnswer":83},"Which differential privacy theories and variants does the survey emphasize?",{"text":84,"@type":76},"It emphasizes recent theoretical developments including Renyi DP and Concentrated DP, along with new mechanisms and techniques and advances in differentially private machine learning.","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"]