{"id":{"repo_id":"msu","oai_identifier":"oai:d.lib.msu.edu:etd_52579"},"canonical_url":"https://search.ndltd.org/etd/msu/oai:d.lib.msu.edu:etd_52579","repository":{"repo_id":"msu","name":"Michigan State University","base_url":"https://d.lib.msu.edu/oai"},"display":{"title":"Knowledge Transfer in Low-rank Matrix Estimation","abstract":"Thesis (Ph.D.)--Michigan State University. Statistics - Doctor of Philosophy, 2025","abstract_html":"Thesis (Ph.D.)--Michigan State University. Statistics - Doctor of Philosophy, 2025","abstract_has_math":false,"creators":["Liu, Dali"],"institution":null,"degree_name":null,"degree_level":null,"degree_discipline":null,"degree_department":null,"school":null,"contributors":["Weng, Haolei H.W","Xiao, Yimin Y.X","Cui, Yuehua Y.C","Xie, Yuying Y.X"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2025,"date_issued":"2025","date_published":"2025","updated_at":"2026-11-18T18:59:09Z","subjects":["Statistics"],"languages":["eng"],"rights":["Attribution-NonCommercial-NoDerivatives 4.0 International"],"rights_urls":[],"identifier_entries":[{"key":"dc:identifier","label":"Identifier","values":["etd:52579","local:Liu_grad.msu_0128D_21535"],"render_values":[{"text":"etd:52579","href":null,"code":true},{"text":"local:Liu_grad.msu_0128D_21535","href":null,"code":true}]}],"person_identifier_entries":[]},"links":{"outbound_url":"https://doi.org/doi:10.25335/f191-4207","outbound_label":"DOI","outbound_source":"dc:identifier","record_url":"https://doi.org/doi:10.25335/f191-4207","record_label":"Repository record","record_kind":"doi","record_source":"dc:identifier","download_url":null,"download_label":null,"download_source":null,"access_status":null,"embargo_until":null,"access_note":null},"source_record":{"url":"https://d.lib.msu.edu/oai?verb=GetRecord&metadataPrefix=oai_dc&identifier=oai%3Ad.lib.msu.edu%3Aetd_52579","prefix":"oai_dc"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Weng, Haolei H.W","Xiao, Yimin Y.X","Cui, Yuehua Y.C","Xie, Yuying Y.X"]},{"key":"dc:creator","label":"Author","values":["Liu, Dali"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2025"]},{"key":"dc:relation","label":"Dc Relation","values":["Electronic Theses & Dissertations"]},{"key":"dc:type","label":"Dc Type","values":["text","Electronic dissertations"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Statistics"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language","label":"Dc Language","values":["eng"]},{"key":"dc:rights","label":"Dc Rights","values":["Attribution-NonCommercial-NoDerivatives 4.0 International"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["etd:52579","local:Liu_grad.msu_0128D_21535","https://doi.org/doi:10.25335/f191-4207"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["Thesis (Ph.D.)--Michigan State University. Statistics - Doctor of Philosophy, 2025","Low-rank modeling plays a pivotal role in signal processing and machine learning, with applications ranging from collaborative filtering, video surveillance, medical imaging, to dimensionality reduction and adaptive filtering.This dissertation explores several questions related to knowledge transfer in low-rank matrix estimation, a central problem in modern statistics with wide ranges of applications. As data from multiple sources become increasingly accessible, new opportunities and challenges arise. We propose methods for leveraging auxiliary information in low-rank matrix estimation, focusing on matrix completion and principal component analysis. All proposed methods are supported by rigorous statistical analysis.We first study the minimax optimal rates in matrix completion. We remove the previously unavoidable logarithmic factor in the convergence rates of several classical estimators, thereby establishing their minimax optimality.Second, we investigate transfer learning for matrix completion. In this problem, data arise from multiple sources and the underlying low-rank structures are similar in nuclear norm. We propose a two-step estimation procedure for the setting where all source datasets are informative. When informative sources are unknown, we develop a data-driven selection procedure.Finally, we study a multi-task learning framework for principal component analysis. We propose a flexible approach for jointly estimating eigenspaces across tasks. Two algorithms are developed under this framework, each achieving a different notion of minimax optimality. The estimators are both adaptive to unknown task similarity structures and robust to arbitrarily distributed outliers.","Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).","Includes bibliographical references."]},{"key":"dc:format","label":"Dc Format","values":["application/pdf","1 online resource (147 p.) : ill."]},{"key":"dc:title","label":"Title","values":["Knowledge Transfer in Low-rank Matrix Estimation"]}]}],"canonical_facts":{"dc:contributor":["Weng, Haolei H.W","Xiao, Yimin Y.X","Cui, Yuehua Y.C","Xie, Yuying Y.X"],"dc:creator":["Liu, Dali"],"dc:date":["2025"],"dc:description":["Thesis (Ph.D.)--Michigan State University. Statistics - Doctor of Philosophy, 2025","Low-rank modeling plays a pivotal role in signal processing and machine learning, with applications ranging from collaborative filtering, video surveillance, medical imaging, to dimensionality reduction and adaptive filtering.This dissertation explores several questions related to knowledge transfer in low-rank matrix estimation, a central problem in modern statistics with wide ranges of applications. As data from multiple sources become increasingly accessible, new opportunities and challenges arise. We propose methods for leveraging auxiliary information in low-rank matrix estimation, focusing on matrix completion and principal component analysis. All proposed methods are supported by rigorous statistical analysis.We first study the minimax optimal rates in matrix completion. We remove the previously unavoidable logarithmic factor in the convergence rates of several classical estimators, thereby establishing their minimax optimality.Second, we investigate transfer learning for matrix completion. In this problem, data arise from multiple sources and the underlying low-rank structures are similar in nuclear norm. We propose a two-step estimation procedure for the setting where all source datasets are informative. When informative sources are unknown, we develop a data-driven selection procedure.Finally, we study a multi-task learning framework for principal component analysis. We propose a flexible approach for jointly estimating eigenspaces across tasks. Two algorithms are developed under this framework, each achieving a different notion of minimax optimality. The estimators are both adaptive to unknown task similarity structures and robust to arbitrarily distributed outliers.","Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).","Includes bibliographical references."],"dc:format":["application/pdf","1 online resource (147 p.) : ill."],"dc:identifier":["etd:52579","local:Liu_grad.msu_0128D_21535","https://doi.org/doi:10.25335/f191-4207"],"dc:language":["eng"],"dc:relation":["Electronic Theses & Dissertations"],"dc:rights":["Attribution-NonCommercial-NoDerivatives 4.0 International"],"dc:subject":["Statistics"],"dc:title":["Knowledge Transfer in Low-rank Matrix Estimation"],"dc:type":["text","Electronic dissertations"]},"updated_at":"2026-11-18T18:59:09Z"}