{"id":{"repo_id":"uiuc","oai_identifier":"oai:www.ideals.illinois.edu:2142/80724"},"canonical_url":"https://search.dev.ndltd.org/etd/uiuc/oai:www.ideals.illinois.edu:2142/80724","repository":{"repo_id":"uiuc","name":"University of Illinois - Urbana-Champaign","base_url":"https://www.ideals.illinois.edu/oai-pmh"},"display":{"title":"Statistical Recursive Estimation Algorithms for Speaker Adaption","abstract":"The major thrust of this thesis is the development of statistical recursive estimation algorithms for speaker adaptation. In this thesis, a unified framework for recursive maximum likelihood estimation and recursive Bayesian learning is first developed and applied to direct hidden Markov model parameter estimation. Then an online Bayesian learning technique is proposed for recursive maximum a posteriori estimation of tree-structured linear regression and affine transformation parameters. This technique has the advantages of accommodating flexible forms of transformation functions as well as prior probability density functions. To balance between model complexity and best goodness of fit to adaptation data, an efficient dynamic programming based pruning algorithm is developed for selecting models using a Bayesian variant of the minimum description length (MDL) principle. Speaker adaptation experiments using a 26-letter English alphabet vocabulary were conducted, and the results confirmed effectiveness of the on-line learning framework.","abstract_html":"The major thrust of this thesis is the development of statistical recursive estimation algorithms for speaker adaptation. In this thesis, a unified framework for recursive maximum likelihood estimation and recursive Bayesian learning is first developed and applied to direct hidden Markov model parameter estimation. Then an online Bayesian learning technique is proposed for recursive maximum a posteriori estimation of tree-structured linear regression and affine transformation parameters. This technique has the advantages of accommodating flexible forms of transformation functions as well as prior probability density functions. To balance between model complexity and best goodness of fit to adaptation data, an efficient dynamic programming based pruning algorithm is developed for selecting models using a Bayesian variant of the minimum description length (MDL) principle. Speaker adaptation experiments using a 26-letter English alphabet vocabulary were conducted, and the results confirmed effectiveness of the on-line learning framework.","abstract_has_math":false,"creators":["Wang, Shaojun"],"institution":"University of Illinois at Urbana-Champaign","degree_name":"Ph.D.","degree_level":"Dissertation","degree_discipline":"Electrical Engineering","degree_department":null,"school":null,"contributors":["Yunxin Zhao"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2015,"date_issued":"2015-09-25T20:07:50Z","date_published":"2015-09-25T20:07:50Z","updated_at":"2026-07-22T22:26:14Z","subjects":["Engineering, Electronics and Electrical"],"languages":["eng"],"rights":[],"rights_urls":[],"identifier_entries":[{"key":"dc:identifier","label":"Identifier","values":["(MiAaPQ)AAI3017245"],"render_values":[{"text":"(MiAaPQ)AAI3017245","href":null,"code":true}]}]},"links":{"outbound_url":"http://hdl.handle.net/2142/80724","outbound_label":"Handle","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Yunxin Zhao"]},{"key":"dc:creator","label":"Author","values":["Wang, Shaojun"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2015-09-25T20:07:50Z","10000-01-01","2001"]},{"key":"dc:type","label":"Dc Type","values":["text"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["Electrical Engineering"]},{"key":"thesis:degree_level","label":"Degree Level","values":["Dissertation"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Ph.D."]},{"key":"thesis:institution_name","label":"Thesis Institution Name","values":["University of Illinois at Urbana-Champaign"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Engineering, Electronics and Electrical"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language","label":"Dc Language","values":["eng"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["http://hdl.handle.net/2142/80724","(MiAaPQ)AAI3017245"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["The major thrust of this thesis is the development of statistical recursive estimation algorithms for speaker adaptation. In this thesis, a unified framework for recursive maximum likelihood estimation and recursive Bayesian learning is first developed and applied to direct hidden Markov model parameter estimation. Then an online Bayesian learning technique is proposed for recursive maximum a posteriori estimation of tree-structured linear regression and affine transformation parameters. This technique has the advantages of accommodating flexible forms of transformation functions as well as prior probability density functions. To balance between model complexity and best goodness of fit to adaptation data, an efficient dynamic programming based pruning algorithm is developed for selecting models using a Bayesian variant of the minimum description length (MDL) principle. Speaker adaptation experiments using a 26-letter English alphabet vocabulary were conducted, and the results confirmed effectiveness of the on-line learning framework.","Made available in DSpace on 2015-09-25T20:07:50Z (GMT). No. of bitstreams: 2 license.txt: 4848 bytes, checksum: 96035ab3f5e1c23cc7138a224ce498bd (MD5) 3017245.pdf: 6132824 bytes, checksum: a825e3dd582f17016e3c030efa5ee2b7 (MD5) Previous issue date: 2001","Embargo set by: Seth Robbins for item 82006 Lift date: Forever Reason: Restricted to the U of I community idenfinitely during batch ingest of legacy ETDs","Restricted to the U of I community idenfinitely during batch ingest of legacy ETDs","U of I Only","127 p.","Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 2001."]},{"key":"dc:title","label":"Title","values":["Statistical Recursive Estimation Algorithms for Speaker Adaption"]}]}],"canonical_facts":{"dc:contributor":["Yunxin Zhao"],"dc:creator":["Wang, Shaojun"],"dc:date":["2015-09-25T20:07:50Z","10000-01-01","2001"],"dc:description":["The major thrust of this thesis is the development of statistical recursive estimation algorithms for speaker adaptation. In this thesis, a unified framework for recursive maximum likelihood estimation and recursive Bayesian learning is first developed and applied to direct hidden Markov model parameter estimation. Then an online Bayesian learning technique is proposed for recursive maximum a posteriori estimation of tree-structured linear regression and affine transformation parameters. This technique has the advantages of accommodating flexible forms of transformation functions as well as prior probability density functions. To balance between model complexity and best goodness of fit to adaptation data, an efficient dynamic programming based pruning algorithm is developed for selecting models using a Bayesian variant of the minimum description length (MDL) principle. Speaker adaptation experiments using a 26-letter English alphabet vocabulary were conducted, and the results confirmed effectiveness of the on-line learning framework.","Made available in DSpace on 2015-09-25T20:07:50Z (GMT). No. of bitstreams: 2 license.txt: 4848 bytes, checksum: 96035ab3f5e1c23cc7138a224ce498bd (MD5) 3017245.pdf: 6132824 bytes, checksum: a825e3dd582f17016e3c030efa5ee2b7 (MD5) Previous issue date: 2001","Embargo set by: Seth Robbins for item 82006 Lift date: Forever Reason: Restricted to the U of I community idenfinitely during batch ingest of legacy ETDs","Restricted to the U of I community idenfinitely during batch ingest of legacy ETDs","U of I Only","127 p.","Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 2001."],"dc:identifier":["http://hdl.handle.net/2142/80724","(MiAaPQ)AAI3017245"],"dc:language":["eng"],"dc:subject":["Engineering, Electronics and Electrical"],"dc:title":["Statistical Recursive Estimation Algorithms for Speaker Adaption"],"dc:type":["text"],"thesis:degree_discipline":["Electrical Engineering"],"thesis:degree_level":["Dissertation"],"thesis:degree_name":["Ph.D."],"thesis:institution_name":["University of Illinois at Urbana-Champaign"]},"updated_at":"2026-07-22T22:26:14Z"}