[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-120594-en":3,"doc-seo-120594-105":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},120594,1099514068365,"Aurelia","https://ap-avatar.wpscdn.com/avatar/10000253d8d9f28188e?_k=1776742907772140068",6,"Technology","Training Quantum Machine Learning Models on Cloud Without Uploading the Data","A run-before-encoding method enables data owners to train quantum machine learning models on quantum cloud platforms without sending raw data, protecting against information leakage. The approach uses linearity of quantum unitary operations to execute parameterized quantum circuits before encoding input data, then computes the cost function on the owner’s local classical computer. It also reduces the encoding bottleneck by lowering circuit depth from O(2^n) to O(n) and relaxes gate precision requirements, while allowing fully classical execution of trained models.","Training quantum machine learning models on cloud without uploading the data  \narXiv :2409 .04602v2 [ quant-ph] 7 Oct 2024  \nGuang Ping He∗  \nSchool of Physics, Sun Yat-sen University, Guangzhou 510275, China  \nBased on the linearity of quantum unitary operations, we propose a method that runs the parameterized quantum circuits before encoding the input data. This enables a dataset owner to train machine learning models on quantum cloud computation platforms, without the risk of leaking the information about the data. It is also capable of encoding a vast amount of data e􀀋ectively at a later time using classical computations, thus saving runtime on quantum computation devices. The trained quantum machine learning models can be run completely on classical computers, meaning the dataset owner does not need to have any quantum hardware, nor even quantum simulators. Moreover, our method mitigates the encoding bottleneck by reducing the required circuit depth from O(2n ) to O (n), and relax the tolerance on the precision of the quantum gates for the encoding. These results demonstrate yet another advantage of quantum and quantum-inspired machine learning models over existing classical neural networks, and broaden the approaches to data security.  \nI. INTRODUCTION  \nData security is rated more and more important nowadays. Individuals need to keep their privacy from misuse and abuse, companies want to protect their intellectual property rights, and governments are concerned about the threat to the national security. Unauthorized transfer of data may be against the law in many places. Some countries also put strict restrictions on exporting data abroad. On the other hand, the rapid development of arti􀀌cial intelligence technology demands a vast amount of data as input, notably in the training of autonomous driving, intelligent healthcare systems and large language models. Meanwhile, not every data owner has very powerful computation devices at home, nor even in his own region. Especially, while quantum machine learning is generally expected to be a potential powerful tool forarti􀀌cial intelligence technology, intermediate and large scale quantum computers are only available via very few cloud platforms in certain countries.  \nTo solve the dilemma between the owner of data and the provider of computational resources, here we propose a run-before-encoding method, which can accomplish the following task. Suppose that Alice owns the dataset and Bob holds the quantum cloud computation platform. Alice can run the quantum circuits on Bob’s platform beforehand, without encoding any data. Based on the output received from Bob, Alice can input her data later and calculate the cost function needed for training machine learning models on her own local classical computer. The 􀀌nal trained models can also be run on Alice’s local classical computer. Since her data has never been sent to Bob’s side, it remains perfectly secure against Bob. This merit makes the method very useful either as a standalone application or as a building block for federated learning [1–3] .  \nThis method is backed by the linearity in quantum  \n􀀃 Electronic address: [hegp@mail.sysu.edu.cn](hegp@mail.sysu.edu.cn)  \nunitary operations, so that it is unavailable for existing classical neural networks where the activation functions of the neurons are nonlinear. Thus, it displays yet another advantage of using quantum or quantum-inspired machine learning models over existing classical counterparts.  \nII. TYPICAL FEATURES OF QUANTUM MACHINE LEARNING MODELS  \nOur method works for quantum circuits with the following features: (1) measurements are performed at the last stage only, while all the rest operations are unitary transformations, and (2) these unitary transformations need to result in real probability amplitudes only. Fortunately, a large portion of the variational quantum circuit (VQC) architecture [4–8] widely used for quantum machine learning today belongs to this category","cbCainYjtPrSuOLs","https://ap.wps.com/l/cbCainYjtPrSuOLs","pdf",448403,1,9,"English","en",105,"# Introduction\n## Data security and computation-resource mismatch\n## Run-before-encoding workflow\n# Typical Features of Quantum Machine Learning Models\n## Circuit constraints: unitary operations and real amplitudes\n## Variational quantum circuit architecture (feature map, ansatz, measurement)\n## Example: RealAmplitudes ansatz and state encoding","[{\"question\":\"What problem does the run-before-encoding method address?\",\"answer\":\"It addresses the dilemma between data owners and cloud providers by enabling training on quantum cloud computation without uploading the dataset, reducing data-leakage risk.\"},{\"question\":\"How does the method let Alice train using only classical computing after interacting with the cloud?\",\"answer\":\"Alice runs parameterized quantum circuits on the cloud before encoding any data, then uses the received outputs to compute the training cost function on her local classical computer and runs the trained models locally.\"},{\"question\":\"What quantum circuit properties does the method require?\",\"answer\":\"The method applies when measurements are performed only at the last stage, and all other operations are unitary transformations that yield real probability amplitudes.\"}]","Training Quantum Machine Learning Models on Cloud Without Uploading the Data | PDF",1785730810,23,{"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},"training-quantum-machine-learning-models-on-cloud-without-uploading-the-data","",{"@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/technology/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/training-quantum-machine-learning-models-on-cloud-without-uploading-the-data/120594/",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},"What problem does the run-before-encoding method address?","Question",{"text":75,"@type":76},"It addresses the dilemma between data owners and cloud providers by enabling training on quantum cloud computation without uploading the dataset, reducing data-leakage risk.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does the method let Alice train using only classical computing after interacting with the cloud?",{"text":80,"@type":76},"Alice runs parameterized quantum circuits on the cloud before encoding any data, then uses the received outputs to compute the training cost function on her local classical computer and runs the trained models locally.",{"name":82,"@type":73,"acceptedAnswer":83},"What quantum circuit properties does the method require?",{"text":84,"@type":76},"The method applies when measurements are performed only at the last stage, and all other operations are unitary transformations that yield real probability amplitudes.","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},[93,97,101,105,110,113,118,123,127,130,134],{"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":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":111,"slug":112},50,"technology",{"id":114,"doc_module":4,"doc_module_name":46,"category_name":115,"show_sort_weight":116,"slug":117},7,"Healthcare",40,"healthcare",{"id":119,"doc_module":4,"doc_module_name":46,"category_name":120,"show_sort_weight":121,"slug":122},8,"Research & Report",30,"research-report",{"id":21,"doc_module":4,"doc_module_name":46,"category_name":124,"show_sort_weight":125,"slug":126},"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"]