{"id":{"repo_id":"buffalo","oai_identifier":"oai:ubir.buffalo.edu:10477/80041"},"canonical_url":"https://search.dev.ndltd.org/etd/buffalo/oai:ubir.buffalo.edu:10477/80041","repository":{"repo_id":"buffalo","name":"Buffalo","base_url":"https://ubir.buffalo.edu/oai/request"},"display":{"title":"High Performance Algorithms for Exact Structure Learning of Bayesian Networks","abstract":"Ph.D.","abstract_html":"Ph.D.","abstract_has_math":false,"creators":["Karan, Subhadeep"],"institution":"State University of New York at Buffalo","degree_name":null,"degree_level":null,"degree_discipline":null,"degree_department":null,"school":null,"contributors":["Zola, Jaroslaw","Computer Science and Engineering"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2019,"date_issued":"2019-07-30T15:12:03Z","date_published":"2019-07-30T15:12:03Z","updated_at":"2026-07-27T19:05:23Z","subjects":["computer science"],"languages":["eng"],"rights":["Users of works found in University at Buffalo Institutional Repository (UBIR) are responsible for identifying and contacting the copyright owner for permission to reuse. University at Buffalo Libraries do not manage rights for copyright-protected works and cannot assist with permissions.","Copyright retained by author."],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"http://hdl.handle.net/10477/80041","outbound_label":"Handle","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Zola, Jaroslaw","Computer Science and Engineering"]},{"key":"dc:creator","label":"Author","values":["Karan, Subhadeep"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2019-07-30T15:12:03Z","2019","2019-05-17 21:28:59"]},{"key":"dc:publisher","label":"Institution","values":["State University of New York at Buffalo"]},{"key":"dc:type","label":"Dc Type","values":["Text","Dissertation"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["computer science"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language","label":"Dc Language","values":["eng"]},{"key":"dc:rights","label":"Dc Rights","values":["Users of works found in University at Buffalo Institutional Repository (UBIR) are responsible for identifying and contacting the copyright owner for permission to reuse. University at Buffalo Libraries do not manage rights for copyright-protected works and cannot assist with permissions.","Copyright retained by author."]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["http://hdl.handle.net/10477/80041"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["Ph.D.","Bayesian networks (BNs) are probabilistic graphical models often used in big data analytics to capture conditional relationships among a set of random variables. They are critical class of models that are capable to represent both as-sociation and causation, and hence are indispensable in building AI systems. Over the years, BNs have been successfully applied in many domains including diagnostic systems, clinical decision support, systems biology, and genomics. However, the problem of learning structure of BNs from data, which is a typical starting point in BNs applications, is known to be NP-hard. To date, both heuristics and exact learning algorithms have been proposed to tackle the problem. While heuristics are widespread in real-life applications, exact algorithms are believed to learn better networks at the much higher computational cost. In this work, we frst show that compared to heuristics the exact structure learning algorithms are delivering measurably better structures in terms of Structural Hamming Distance, and enable more accurate inferences, measured through the Kullback-Leiber divergence between the actual and inferred probability distributions. Then, we propose high performance strategies to address computational challenges of exact structure learning algorithms."]},{"key":"dc:format","label":"Dc Format","values":["application/pdf"]},{"key":"dc:title","label":"Title","values":["High Performance Algorithms for Exact Structure Learning of Bayesian Networks"]}]}],"canonical_facts":{"dc:contributor":["Zola, Jaroslaw","Computer Science and Engineering"],"dc:creator":["Karan, Subhadeep"],"dc:date":["2019-07-30T15:12:03Z","2019","2019-05-17 21:28:59"],"dc:description":["Ph.D.","Bayesian networks (BNs) are probabilistic graphical models often used in big data analytics to capture conditional relationships among a set of random variables. They are critical class of models that are capable to represent both as-sociation and causation, and hence are indispensable in building AI systems. Over the years, BNs have been successfully applied in many domains including diagnostic systems, clinical decision support, systems biology, and genomics. However, the problem of learning structure of BNs from data, which is a typical starting point in BNs applications, is known to be NP-hard. To date, both heuristics and exact learning algorithms have been proposed to tackle the problem. While heuristics are widespread in real-life applications, exact algorithms are believed to learn better networks at the much higher computational cost. In this work, we frst show that compared to heuristics the exact structure learning algorithms are delivering measurably better structures in terms of Structural Hamming Distance, and enable more accurate inferences, measured through the Kullback-Leiber divergence between the actual and inferred probability distributions. Then, we propose high performance strategies to address computational challenges of exact structure learning algorithms."],"dc:format":["application/pdf"],"dc:identifier":["http://hdl.handle.net/10477/80041"],"dc:language":["eng"],"dc:publisher":["State University of New York at Buffalo"],"dc:rights":["Users of works found in University at Buffalo Institutional Repository (UBIR) are responsible for identifying and contacting the copyright owner for permission to reuse. University at Buffalo Libraries do not manage rights for copyright-protected works and cannot assist with permissions.","Copyright retained by author."],"dc:subject":["computer science"],"dc:title":["High Performance Algorithms for Exact Structure Learning of Bayesian Networks"],"dc:type":["Text","Dissertation"]},"updated_at":"2026-07-27T19:05:23Z"}