{"id":{"repo_id":"msu","oai_identifier":"oai:d.lib.msu.edu:etd_52906"},"canonical_url":"https://search.ndltd.org/etd/msu/oai:d.lib.msu.edu:etd_52906","repository":{"repo_id":"msu","name":"Michigan State University","base_url":"https://d.lib.msu.edu/oai"},"display":{"title":"REGULARITY IN OPTIMALITY : A MULTI-DOMAIN STUDY OF STRUCTURED MULTI-OBJECTIVE OPTIMIZATION","abstract":"Thesis (Ph.D.)--Michigan State University. Computer Science - Doctor of Philosophy, 2025","abstract_html":"Thesis (Ph.D.)--Michigan State University. Computer Science - Doctor of Philosophy, 2025","abstract_has_math":false,"creators":["Guha, Ritam"],"institution":null,"degree_name":null,"degree_level":null,"degree_discipline":null,"degree_department":null,"school":null,"contributors":["Deb, Kalyanmoy KD","Banzhaf, Wolfgang WB","Dolson, Emily ED","Kordjamshidi, Parisa PK","Mckendrick, Ryan RM"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2025,"date_issued":"2025","date_published":"2025","updated_at":"2026-11-18T18:59:31Z","subjects":["Computer science","Artificial intelligence"],"languages":["eng"],"rights":["In Copyright"],"rights_urls":[],"identifier_entries":[{"key":"dc:identifier","label":"Identifier","values":["etd:52906","local:Guha_grad.msu_0128D_21900"],"render_values":[{"text":"etd:52906","href":null,"code":true},{"text":"local:Guha_grad.msu_0128D_21900","href":null,"code":true}]}],"person_identifier_entries":[]},"links":{"outbound_url":"https://doi.org/doi:10.25335/ecfy-7385","outbound_label":"DOI","outbound_source":"dc:identifier","record_url":"https://doi.org/doi:10.25335/ecfy-7385","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_52906","prefix":"oai_dc"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Deb, Kalyanmoy KD","Banzhaf, Wolfgang WB","Dolson, Emily ED","Kordjamshidi, Parisa PK","Mckendrick, Ryan RM"]},{"key":"dc:creator","label":"Author","values":["Guha, Ritam"]}]},{"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":["Computer science","Artificial intelligence"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language","label":"Dc Language","values":["eng"]},{"key":"dc:rights","label":"Dc Rights","values":["In Copyright"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["etd:52906","local:Guha_grad.msu_0128D_21900","https://doi.org/doi:10.25335/ecfy-7385"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["Thesis (Ph.D.)--Michigan State University. Computer Science - Doctor of Philosophy, 2025","Multi-objective optimization problems give rise to a set of Pareto-optimal solutions, each of which makes a trade-off among the objectives. When multiple Pareto-optimal solutions are to be implemented for different applications as platform-based solutions, a solution principle common to them is highly desired for easier understanding, implementation, and management purposes. This report proposes a systematic search methodology that deviates from finding Pareto-optimal solutions but finds a set of near-Pareto-optimal solutions sharing common principles of a desired structure and still possessing a trade-off among objectives. After proposing the regular evolutionary multi-objective optimization (RegEMO) algorithm, the report first demonstrates its working principle on a number of constrained and unconstrained multi-objective test problems. Thereafter, the practical significance of the proposed approach has been validated through the application to a number of engineering design problems. Searching for a set of solutions with common principles of desire, rather than theoretical Pareto-optimal solutions without any common structure, is a practically meaningful task and this work should encourage more such practice-oriented developments of EMO in the near future.While the framework is broadly defined to accommodate different notions of regularity across problem domains, the initial focus is on continuous numerical optimization problems where regularity is expressed as relationships among design variables. The concept is then extended to the domain of Neural Architecture Search (NAS) through a framework called Regularity-based Multi-objective Neural Architecture Search (RegMONAS). Here, the objective is not only to balance conflicting metrics like accuracy and latency but also to uncover regular design patterns in network architectures that promote hardware-friendliness and reusability. RegMONAS showcases the adaptability of the regularity concept to complex deep learning model design, where interpretable and modular architectures are often more valuable than marginal performance gains.Finally, the regularity-based framework is applied to a complex multi-objective wargame strategy optimization problem, developed in collaboration with Northrop Grumman Corporation. This problem involves two competing agents, offense and defense, whose strategies are highly interdependent. An optimal strategy for one agent depends on the anticipated moves of the other. To address this, a multi-objective competitive co-evolutionary optimization framework was employed. As the decision-making integrated wargame strategy optimization progresses, certain variables become fixed to ensure a feasible transition to new strategies after committing to the previous ones. This variable fixing process represents the definition of regularity in this context. This collaboration between RegEMO and co-evolution highlights the versatility of the regularity-based optimization framework and demonstrates its applicability to complex, real-world problems.In summary, this report showcases how the concept of regularity can bridge the gap between theoretical optimization and practical implementation, offering a robust tool for solving diverse real-life challenges.","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 (143 p.) : ill."]},{"key":"dc:title","label":"Title","values":["REGULARITY IN OPTIMALITY : A MULTI-DOMAIN STUDY OF STRUCTURED MULTI-OBJECTIVE OPTIMIZATION"]}]}],"canonical_facts":{"dc:contributor":["Deb, Kalyanmoy KD","Banzhaf, Wolfgang WB","Dolson, Emily ED","Kordjamshidi, Parisa PK","Mckendrick, Ryan RM"],"dc:creator":["Guha, Ritam"],"dc:date":["2025"],"dc:description":["Thesis (Ph.D.)--Michigan State University. Computer Science - Doctor of Philosophy, 2025","Multi-objective optimization problems give rise to a set of Pareto-optimal solutions, each of which makes a trade-off among the objectives. When multiple Pareto-optimal solutions are to be implemented for different applications as platform-based solutions, a solution principle common to them is highly desired for easier understanding, implementation, and management purposes. This report proposes a systematic search methodology that deviates from finding Pareto-optimal solutions but finds a set of near-Pareto-optimal solutions sharing common principles of a desired structure and still possessing a trade-off among objectives. After proposing the regular evolutionary multi-objective optimization (RegEMO) algorithm, the report first demonstrates its working principle on a number of constrained and unconstrained multi-objective test problems. Thereafter, the practical significance of the proposed approach has been validated through the application to a number of engineering design problems. Searching for a set of solutions with common principles of desire, rather than theoretical Pareto-optimal solutions without any common structure, is a practically meaningful task and this work should encourage more such practice-oriented developments of EMO in the near future.While the framework is broadly defined to accommodate different notions of regularity across problem domains, the initial focus is on continuous numerical optimization problems where regularity is expressed as relationships among design variables. The concept is then extended to the domain of Neural Architecture Search (NAS) through a framework called Regularity-based Multi-objective Neural Architecture Search (RegMONAS). Here, the objective is not only to balance conflicting metrics like accuracy and latency but also to uncover regular design patterns in network architectures that promote hardware-friendliness and reusability. RegMONAS showcases the adaptability of the regularity concept to complex deep learning model design, where interpretable and modular architectures are often more valuable than marginal performance gains.Finally, the regularity-based framework is applied to a complex multi-objective wargame strategy optimization problem, developed in collaboration with Northrop Grumman Corporation. This problem involves two competing agents, offense and defense, whose strategies are highly interdependent. An optimal strategy for one agent depends on the anticipated moves of the other. To address this, a multi-objective competitive co-evolutionary optimization framework was employed. As the decision-making integrated wargame strategy optimization progresses, certain variables become fixed to ensure a feasible transition to new strategies after committing to the previous ones. This variable fixing process represents the definition of regularity in this context. This collaboration between RegEMO and co-evolution highlights the versatility of the regularity-based optimization framework and demonstrates its applicability to complex, real-world problems.In summary, this report showcases how the concept of regularity can bridge the gap between theoretical optimization and practical implementation, offering a robust tool for solving diverse real-life challenges.","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 (143 p.) : ill."],"dc:identifier":["etd:52906","local:Guha_grad.msu_0128D_21900","https://doi.org/doi:10.25335/ecfy-7385"],"dc:language":["eng"],"dc:relation":["Electronic Theses & Dissertations"],"dc:rights":["In Copyright"],"dc:subject":["Computer science","Artificial intelligence"],"dc:title":["REGULARITY IN OPTIMALITY : A MULTI-DOMAIN STUDY OF STRUCTURED MULTI-OBJECTIVE OPTIMIZATION"],"dc:type":["text","Electronic dissertations"]},"updated_at":"2026-11-18T18:59:31Z"}