[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-120009-en":3,"doc-seo-120009-105":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},120009,137441390410,"Hazel","https://ap-avatar.wpscdn.com/avatar/2000252f4ab5702993?_k=1776741390130283984",8,"Research & Report","Harnessing Inherent Noises for Privacy Preservation in Quantum Machine Learning","Quantum computing accelerates complex computations and large-scale data processing, yet privacy leakage in quantum machine learning (QML) can expose sensitive information. Differential privacy (DP) uses artificial noise but remains insufficiently explored for QML. This paper leverages inherent shot noise and incoherent noise in NISQ devices to protect QML models for binary classification. It proves that gradients follow a Gaussian distribution and derives variance bounds, supporting DP guarantees via simulations that reach target privacy levels by varying circuit repetitions.","Harnessing Inherent Noises for Privacy Preservation in Quantum Machine Learning  \nKeyi Ju 1 , Xiaoqi Qin 1 , Hui Zhong2 , Xinyue Zhang3 , Miao Pan2 , Baoling Liu 1  \n1 State Key Laboratory of Networking and Switching Technology,  \nBeijing University of Posts and Telecommunications, Beijing, China, 100876  \n2Department of Electrical and Computer Engineering, University of Houston, Houston, TX, 77204  \n3Department of Computer Science, Kennesaw State University, Marietta, GA, 30060  \narXiv :2312 . 11126v2 [ quant-ph] 7 Mar 2024  \nAbstract—Quantum computing revolutionizes the way of solving complex problems and handling vast datasets, which shows great potential to accelerate the machine learning process. However, data leakage in quantum machine learning (QML) may present privacy risks. Although differential privacy (DP), which protects privacy through the injection of artificial noise, is a well-established approach, its application in the QML domain remains under-explored. In this paper, we propose to harness inherent quantum noises to protect data privacy in QML. Especially, considering the Noisy Intermediate-Scale Quantum (NISQ) devices, we leverage the unavoidable shot noise and incoherent noise in quantum computing to preserve the privacy of QML models for binary classification. We mathematically analyze that the gradient of quantum circuit parameters in QML satisfies a Gaussian distribution, and derive the upper and lower boundson its variance, which can potentially provide the DP guarantee. Through simulations, we show that a target privacy protection level can be achieved by running the quantum circuit a different number of times.  \nIndex Terms—Quantum machine learning, Quantum differential privacy, Shot noise, Incoherent noise  \nI. INTRODUCTION  \nQuantum computing, which exploits principles of quantum mechanics such as superposition and entanglement, offers a great potential for significantly faster complex calculations than classical computers. In the Noisy IntermediateScale Quantum (NISQ) era, variational quantum algorithms (VQAs) are considered the best for quantum computing, which heavily rely on variational quantum circuits (VQCs) . VQCs are parameterized quantum circuits designed to be optimized iteratively to minimize specific cost functions, enabling the exploration of solution spaces and the identification of optimal quantum states. Based on the advance of these technologies, quantum computing is becoming a powerful alternative to conduct computing intensive machine learning or optimization tasks [1, 2] .  \nQuantum machine learning (QML) has demonstrated advantages over classical machine learning (CML) in dealing with large datasets and complex problems [3] . Due to its high computing efficiency, QML has the potential to be used in various areas such as nanoparticle synthesis and biomedical domain. Nevertheless, the datasets used in these applications may be highly sensitive, e.g., CT scanned images of COVID- 19 patients [4] . Therefore, similar to data privacy concerns  \nin CML, privacy preservation is crucial in QML to safeguard sensitive data from potential breaches and misuse.  \nIn CML scenarios, we integrate differential privacy (DP) techniques to protect sensitive data and mitigate privacy risks during data processing and analysis. DP is a concept in data science and statistics that aims to provide a means of statistical database privacy protection proposed by [5] . It offers a way to maximize the accuracy of queries from statistical databases while minimizing the chances of identifying information about specific individuals within the database. By introducing controlled amounts of statistical noise to the data, DP ensures that the presence or absence of specific records does not significantly affect the results, thereby safeguarding individual privacy. One of the most popular DP definitions is (ϵ,δ)-DP, where ϵ represents the maximum allowed change in output due to the addition or removal of an individual","cbCaioSl7SeHUOf5","https://ap.wps.com/l/cbCaioSl7SeHUOf5","pdf",499313,1,6,"English","en",105,"# Introduction\n## Differential privacy in machine learning\n## Extending DP to quantum computing\n## Inherent shot noise and incoherent noise in QML\n## Research approach and quantum error mitigation","[{\"question\":\"为什么量子机器学习需要隐私保护？\",\"answer\":\"QML可能处理高度敏感的数据，类似于经典机器学习中的隐私风险，可能导致数据泄露与滥用。因此需要隐私保护机制。\"},{\"question\":\"论文如何利用量子电路的固有噪声来实现隐私保护？\",\"answer\":\"在NISQ设备背景下，论文利用不可避免的shot noise与非相干噪声作为隐私保护噪声来源，替代或补充对梯度注入的人工噪声思路。\"},{\"question\":\"隐私保证是如何与训练过程建立联系的？\",\"answer\":\"论文从理论上分析QML中量子电路参数梯度的统计性质，证明其满足高斯分布并给出方差的上下界，从而可用于支持(ε,δ)-DP的隐私保证。\"}]","Harnessing Inherent Noises for Privacy Preservation in Quantum Machine Learning | PDF",1785727690,15,{"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},"harnessing-inherent-noises-for-privacy-preservation-in-quantum-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/harnessing-inherent-noises-for-privacy-preservation-in-quantum-machine-learning/120009/",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","2026-08-03",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},"为什么量子机器学习需要隐私保护？","Question",{"text":75,"@type":76},"QML可能处理高度敏感的数据，类似于经典机器学习中的隐私风险，可能导致数据泄露与滥用。因此需要隐私保护机制。","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"论文如何利用量子电路的固有噪声来实现隐私保护？",{"text":80,"@type":76},"在NISQ设备背景下，论文利用不可避免的shot noise与非相干噪声作为隐私保护噪声来源，替代或补充对梯度注入的人工噪声思路。",{"name":82,"@type":73,"acceptedAnswer":83},"隐私保证是如何与训练过程建立联系的？",{"text":84,"@type":76},"论文从理论上分析QML中量子电路参数梯度的统计性质，证明其满足高斯分布并给出方差的上下界，从而可用于支持(ε,δ)-DP的隐私保证。","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},[93,97,101,105,110,114,119,122,127,130,134],{"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":21,"doc_module":4,"doc_module_name":46,"category_name":111,"show_sort_weight":112,"slug":113},"Technology",50,"technology",{"id":115,"doc_module":4,"doc_module_name":46,"category_name":116,"show_sort_weight":117,"slug":118},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":120,"slug":121},30,"research-report",{"id":123,"doc_module":4,"doc_module_name":46,"category_name":124,"show_sort_weight":125,"slug":126},9,"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"]