{"id":{"repo_id":"odu","oai_identifier":"oai:digitalcommons.odu.edu:computerscience_etds-1185"},"canonical_url":"https://search.ndltd.org/etd/odu/oai:digitalcommons.odu.edu:computerscience_etds-1185","repository":{"repo_id":"odu","name":"Old Dominion University","base_url":"https://digitalcommons.odu.edu/do/oai/"},"display":{"title":"Symbolic Regression for Data-Driven Equation Discovery: A Physics-Informed Approach","abstract":"Symbolic Regression (SR) is a cutting-edge machine learning technique that discovers mathematical expressions representing the underlying patterns in data. Unlike traditional regression, SR explores a wide range of mathematical models, allowing for flexible and interpretable solutions. We utilize PySR, a highly customizable symbolic regression package, which combines genetic programming and modern optimization methods to efficiently search for interpretable equations. PySR balances model complexity with performance by penalizing overly complex expressions while optimizing accuracy. In this thesis, we apply Physics-informed Symbolic Regression through PySR to model the x and t dependence of the flavor isovector combination Hu − d(x,t,ζ ,Q2 ) at ζ = 0 and Q2 = 4GeV2 . Our PySR models are trained on Generalized Parton Distribution (GPD) pseudo data, sourced from both Lattice QCD and contemporary models such as GGL. By incorporating physics constraints into the symbolic regression process, our models satisfy physical principles relevant to GPDs while achieving low mean-squared error. A custom loss function was employed, encouraging models that both fit the data well and adhered to physical constraints. We compare the performance of PySR-derived GPDs against traditional Regge parameterizations and Neural Network-based GPDs.This research demonstrates the power of PySR for extracting interpretable expressions from complex physics data, offering a novel tool for scientific modeling that merges data-driven techniques with domain-specific knowledge.","abstract_html":"Symbolic Regression (SR) is a cutting-edge machine learning technique that discovers mathematical expressions representing the underlying patterns in data. Unlike traditional regression, SR explores a wide range of mathematical models, allowing for flexible and interpretable solutions. We utilize PySR, a highly customizable symbolic regression package, which combines genetic programming and modern optimization methods to efficiently search for interpretable equations. PySR balances model complexity with performance by penalizing overly complex expressions while optimizing accuracy. In this thesis, we apply Physics-informed Symbolic Regression through PySR to model the x and t dependence of the flavor isovector combination Hu − d(x,t,ζ ,Q2 ) at ζ = 0 and Q2 = 4GeV2 . Our PySR models are trained on Generalized Parton Distribution (GPD) pseudo data, sourced from both Lattice QCD and contemporary models such as GGL. By incorporating physics constraints into the symbolic regression process, our models satisfy physical principles relevant to GPDs while achieving low mean-squared error. A custom loss function was employed, encouraging models that both fit the data well and adhered to physical constraints. We compare the performance of PySR-derived GPDs against traditional Regge parameterizations and Neural Network-based GPDs.This research demonstrates the power of PySR for extracting interpretable expressions from complex physics data, offering a novel tool for scientific modeling that merges data-driven techniques with domain-specific knowledge.","abstract_has_math":false,"creators":["Singireddy, Anusha Reddy"],"institution":null,"degree_name":"Master of Science (MS)","degree_level":"Thesis","degree_discipline":"Computer Science","degree_department":null,"school":null,"contributors":["Yaohang Li","Ravi Mukkamala","Douglas Adams"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":null,"date_issued":"","date_published":null,"updated_at":"2026-11-19T08:32:56Z","subjects":["Computer Sciences","Physics"],"languages":[],"rights":["<p>In Copyright. URI: <a href=\"http://rightsstatements.org/vocab/InC/1.0/\">http://rightsstatements.org/vocab/InC/1.0/</a> This Item is protected by copyright and/or related rights. You are free to use this Item in any way that is permitted by the copyright and related rights legislation that applies to your use. For other uses you need to obtain permission from the rights-holder(s).</p>"],"rights_urls":[],"identifier_entries":[{"key":"dc:identifier","label":"Identifier","values":["9798302855473"],"render_values":[{"text":"9798302855473","href":null,"code":true}]}],"person_identifier_entries":[]},"links":{"outbound_url":"https://digitalcommons.odu.edu/computerscience_etds/185","outbound_label":"Repository record","outbound_source":"dc:identifier","record_url":"https://digitalcommons.odu.edu/computerscience_etds/185","record_label":"Repository record","record_kind":"record","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://digitalcommons.odu.edu/do/oai/?verb=GetRecord&metadataPrefix=oai_etdms&identifier=oai%3Adigitalcommons.odu.edu%3Acomputerscience_etds-1185","prefix":"oai_etdms"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Yaohang Li","Ravi Mukkamala","Douglas Adams"]},{"key":"dc:creator","label":"Author","values":["Singireddy, Anusha Reddy"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.available","label":"Dc Date Available","values":["2025-02-11T08:00:00Z"]},{"key":"dc:date.created","label":"Dc Date Created","values":["2024-10-01T07:00:00Z"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["Computer Science"]},{"key":"thesis:degree_level","label":"Degree Level","values":["Thesis"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Master of Science (MS)"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Computer Sciences","Physics"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:rights","label":"Dc Rights","values":["<p>In Copyright. URI: <a href=\"http://rightsstatements.org/vocab/InC/1.0/\">http://rightsstatements.org/vocab/InC/1.0/</a> This Item is protected by copyright and/or related rights. You are free to use this Item in any way that is permitted by the copyright and related rights legislation that applies to your use. For other uses you need to obtain permission from the rights-holder(s).</p>"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["9798302855473","https://digitalcommons.odu.edu/computerscience_etds/185"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["Symbolic Regression (SR) is a cutting-edge machine learning technique that discovers mathematical expressions representing the underlying patterns in data. Unlike traditional regression, SR explores a wide range of mathematical models, allowing for flexible and interpretable solutions. We utilize PySR, a highly customizable symbolic regression package, which combines genetic programming and modern optimization methods to efficiently search for interpretable equations. PySR balances model complexity with performance by penalizing overly complex expressions while optimizing accuracy. In this thesis, we apply Physics-informed Symbolic Regression through PySR to model the x and t dependence of the flavor isovector combination Hu − d(x,t,ζ ,Q2 ) at ζ = 0 and Q2 = 4GeV2 . Our PySR models are trained on Generalized Parton Distribution (GPD) pseudo data, sourced from both Lattice QCD and contemporary models such as GGL. By incorporating physics constraints into the symbolic regression process, our models satisfy physical principles relevant to GPDs while achieving low mean-squared error. A custom loss function was employed, encouraging models that both fit the data well and adhered to physical constraints. We compare the performance of PySR-derived GPDs against traditional Regge parameterizations and Neural Network-based GPDs.This research demonstrates the power of PySR for extracting interpretable expressions from complex physics data, offering a novel tool for scientific modeling that merges data-driven techniques with domain-specific knowledge."]},{"key":"dc:title","label":"Title","values":["Symbolic Regression for Data-Driven Equation Discovery: A Physics-Informed Approach"]}]}],"canonical_facts":{"dc:contributor":["Yaohang Li","Ravi Mukkamala","Douglas Adams"],"dc:creator":["Singireddy, Anusha Reddy"],"dc:date.available":["2025-02-11T08:00:00Z"],"dc:date.created":["2024-10-01T07:00:00Z"],"dc:description.abstract":["Symbolic Regression (SR) is a cutting-edge machine learning technique that discovers mathematical expressions representing the underlying patterns in data. Unlike traditional regression, SR explores a wide range of mathematical models, allowing for flexible and interpretable solutions. We utilize PySR, a highly customizable symbolic regression package, which combines genetic programming and modern optimization methods to efficiently search for interpretable equations. PySR balances model complexity with performance by penalizing overly complex expressions while optimizing accuracy. In this thesis, we apply Physics-informed Symbolic Regression through PySR to model the x and t dependence of the flavor isovector combination Hu − d(x,t,ζ ,Q2 ) at ζ = 0 and Q2 = 4GeV2 . Our PySR models are trained on Generalized Parton Distribution (GPD) pseudo data, sourced from both Lattice QCD and contemporary models such as GGL. By incorporating physics constraints into the symbolic regression process, our models satisfy physical principles relevant to GPDs while achieving low mean-squared error. A custom loss function was employed, encouraging models that both fit the data well and adhered to physical constraints. We compare the performance of PySR-derived GPDs against traditional Regge parameterizations and Neural Network-based GPDs.This research demonstrates the power of PySR for extracting interpretable expressions from complex physics data, offering a novel tool for scientific modeling that merges data-driven techniques with domain-specific knowledge."],"dc:identifier":["9798302855473","https://digitalcommons.odu.edu/computerscience_etds/185"],"dc:rights":["<p>In Copyright. URI: <a href=\"http://rightsstatements.org/vocab/InC/1.0/\">http://rightsstatements.org/vocab/InC/1.0/</a> This Item is protected by copyright and/or related rights. You are free to use this Item in any way that is permitted by the copyright and related rights legislation that applies to your use. For other uses you need to obtain permission from the rights-holder(s).</p>"],"dc:subject":["Computer Sciences","Physics"],"dc:title":["Symbolic Regression for Data-Driven Equation Discovery: A Physics-Informed Approach"],"thesis:degree_discipline":["Computer Science"],"thesis:degree_level":["Thesis"],"thesis:degree_name":["Master of Science (MS)"]},"updated_at":"2026-11-19T08:32:56Z"}