[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-127556-en":3,"doc-seo-127556-105":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},127556,687207020761,"Patrick","https://ap-avatar.wpscdn.com/davatar_155a257f0dc6eb9ab79c44ca47cae57d",8,"Research & Report","An Improved Classical Singular Value Transformation for Quantum Machine Learning - arXiv Preprint","Quantum machine learning (QML) aims to achieve quantum speedups for machine learning and data analysis, but such advantages require classical algorithms to be asymptotically slower. This work studies quantum-classical gaps using the quantum singular value transformation (QSVT) framework, which unifies major quantum speedup techniques. The paper challenges QSVT-based low-rank applications by presenting a classical algorithm that matches the QSVT performance up to a small polynomial overhead. It leverages Clenshaw recurrence for matrix polynomials and sketching to simulate QSVT classically, introducing new sketching, stability, and coefficient-bounding tools.","An Improved Classical Singular Value Transformation for Quantum Machine Learning  \nAinesh Bakshi[ainesh@mit.edu](ainesh@mit.edu)[ ](ainesh@mit.edu)MIT  \nEwin Tang [ewint@cs.washington.edu](ewint@cs.washington.edu)[ ](ewint@cs.washington.edu)University of Washington  \narXiv :2303 .01492v3 [ quant-ph] 3 Aug 2023  \nAbstract  \nThe field of quantum machine learning (QML) produces many proposals for attaining quantum speedups for tasks in machine learning and data analysis. Such speedups can only manifest if classical algorithms for these tasks perform significantly slower than quantum ones. We study quantum-classical gaps in QML through the quantum singular value transformation (QSVT) framework. QSVT, introduced by Gilyén, Su, Low and Wiebe [GSLW19], unifies all major types of quantum speedup [MRTC21]; in particular, a wide variety of QML proposals are applications of QSVT on low-rank classical data. We challenge these proposals by providing a classical algorithm that matches the performance of QSVT in this regime up to a small polynomial overhead.  \nWe show that, given a matrix A ∈ Cm ×n, a vector b ∈ Cn, a bounded degree-d polynomial p, and linear-time pre-processing, we can output a description of a vector v such -ahapssec(aas)bl apsi∥ot∥r[nCTy,,(s∥A22afim∥t,(ε2hiinth4eqtimag)un) timiree pitu-.po(∥p6F/ngtryr()∥n,e nmec(Aonen∥F/pdatri(iAcslss))ea[lT.gaistth9;anmCatfoLingr quLTWr]u, -ritspHilreiaeu[ntCGlatpLio[nWC;ialGLTsSTW, e]],  \nOur key insight is to combine the Clenshaw recurrence, an iterative method for computing matrix polynomials, with sketching techniques to simulate QSVT classically. We introduce several new classical techniques in this work, including (a) a non-oblivious matrix sketch for approximately preserving bi-linear forms,(b) a new stability analysis for the Clenshaw recurrence, and (c) a new technique to bound arithmetic progressions of the coefficients appearing in the Chebyshev series expansion of bounded functions, each of which may be of independent interest.  \nContents  \n1 Introduction 1  \n1.1 Results ........................................... 2  \n1.2 Applications to Dequantizing QML .......................... 4  \n2 Technical Overview 5  \n3 Discussion 12  \n4 Preliminaries 14  \n4.1 Linear Algebra ....................................... 14  \n4.2 Polynomials and the Chebyshev Basis ......................... 15  \n4.3 Sampling and Query Access ............................... 16  \n5 Extending the Sketching Toolkit 19  \n5.1 The Bi-Linear Entry-wise Sampling Transform .................... 19  \n5.2 Approximate Matrix Products from ℓ22 Importance Sampling ............ 20  \n6 Sums of Chebyshev Coefficients 23  \n7 Properties of the Clenshaw Recursion 29  \n7.1 Deriving the Clenshaw Recursions ........................... 29  \n7.2 Evaluating Even and Odd Polynomials ........................ 30  \n8 Stability of the Scalar Clenshaw Recursion 32  \n8.1 Analyzing Error Propagation .............................. 32  \n8.2 Bounding the Iterates of the Clenshaw Recurrence .................. 35  \n9 Computing Matrix Polynomials 36  \n9.1 Computing Odd Matrix Polynomials .......................... 37  \n9.2 Generalizing to Even Polynomials ........................... 44  \n9.3 Bounding Iterates of Singular Value Transformation ................. 49  \n10 Dequantizing QML Algorithms 51  \n10.1 Recommendation Systems ................................ 51  \n10.2 Linear Regression ..................................... 52  \n10.3 Hamiltonian Simulation ................................. 53  \nAcknowledgments 54  \nReferences 54  \n1 Introduction  \nQuantum machine learning (QML) has rapidly developed into an active field of study with numerous proposals for speeding up machine learning tasks with quantum computers [DW20; Cil+18] . These proposals include quantum algorithms for basic tasks in machine learning, including regression [HHL09], perceptron learning [KWS16], support vector machines [RML14], recommendation systems [KP17], and semi-definite pr","cbCail9hsASbc7sG","https://ap.wps.com/l/cbCail9hsASbc7sG","pdf",795947,1,62,"English","en",105,"# Abstract\n# Introduction\n## Results\n## Applications to Dequantizing QML\n# Technical Overview\n# Discussion\n# Preliminaries\n## Linear Algebra\n## Polynomials and the Chebyshev Basis\n## Sampling and Query Access\n# Extending the Sketching Toolkit\n## The Bi-Linear Entry-wise Sampling Transform\n## Approximate Matrix Products from ℓ22 Importance Sampling\n# Sums of Chebyshev Coefficients\n# Properties of the Clenshaw Recursion\n## Deriving the Clenshaw Recursions\n## Evaluating Even and Odd Polynomials\n# Stability of the Scalar Clenshaw Recursion\n## Analyzing Error Propagation\n## Bounding the Iterates of the Clenshaw Recurrence\n# Computing Matrix Polynomials\n## Computing Odd Matrix Polynomials\n## Generalizing to Even Polynomials\n## Bounding Iterates of Singular Value Transformation\n# Dequantizing QML Algorithms\n## Recommendation Systems\n## Linear Regression\n## Hamiltonian Simulation\n# Acknowledgments\n# References","[{\"question\":\"What gap between quantum and classical algorithms does the paper study in QML?\",\"answer\":\"It studies quantum-classical gaps in quantum machine learning through the quantum singular value transformation (QSVT) framework, focusing on regimes where QSVT applies to low-rank classical data.\"},{\"question\":\"How does the proposed classical algorithm relate to QSVT performance?\",\"answer\":\"It provides a classical algorithm that matches the QSVT performance in the targeted regime up to a small polynomial overhead.\"},{\"question\":\"What technical ideas enable the classical simulation of QSVT?\",\"answer\":\"The work combines Clenshaw recurrence for computing matrix polynomials with sketching techniques for classical simulation, including a non-oblivious matrix sketch for preserving bilinear forms and new stability and coefficient-bounding analyses.\"}]","An Improved Classical Singular Value Transformation for Quantum Machine Learning - arXiv Preprint | PDF",1785939941,156,{"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},"an-improved-classical-singular-value-transformation-for-quantum-machine-learning-arxiv-preprint","",{"@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/an-improved-classical-singular-value-transformation-for-quantum-machine-learning-arxiv-preprint/127556/",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","2026-08-23","2026-08-05",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 gap between quantum and classical algorithms does the paper study in QML?","Question",{"text":76,"@type":77},"It studies quantum-classical gaps in quantum machine learning through the quantum singular value transformation (QSVT) framework, focusing on regimes where QSVT applies to low-rank classical data.","Answer",{"name":79,"@type":74,"acceptedAnswer":80},"How does the proposed classical algorithm relate to QSVT performance?",{"text":81,"@type":77},"It provides a classical algorithm that matches the QSVT performance in the targeted regime up to a small polynomial overhead.",{"name":83,"@type":74,"acceptedAnswer":84},"What technical ideas enable the classical simulation of QSVT?",{"text":85,"@type":77},"The work combines Clenshaw recurrence for computing matrix polynomials with sketching techniques for classical simulation, including a non-oblivious matrix sketch for preserving bilinear forms and new stability and coefficient-bounding analyses.","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"]