{"id":{"repo_id":"vilnius","oai_identifier":"oai:vu.lt:elaba:210578055"},"canonical_url":"https://search.dev.ndltd.org/etd/vilnius/oai:vu.lt:elaba:210578055","repository":{"repo_id":"vilnius","name":"Vilnius University","base_url":"https://epublications.vu.lt/oai"},"display":{"title":"Bazinių mašininio mokymo modelių taikymas kvantinės chemijos duomenų analizei /","abstract":"Understanding and exploring the chemical environment, particularly the dynamics of large molecular structures, is crucial in molecular chemistry. However, the computational resources required for solving such complex systems often necessitate the use of approximate calculation methods, which can compromise result accuracy. To address it, this research explores the application of foundation machine learning models to analyze quantum chemistry data, specifically focusing on predicting molecular energy states accurately without sacrificing computational efficiency. The main problem addressed by this research is the accurate determination of a molecule's ground state energy from its Cartesian coordinates, a task complicated by the need for substantial computational power and precision. The primary objective is to assess the potential application of foundation machine learning models for quantum chemistry problems. The main tasks that are involved in achieving the said objective are understanding the fonudation models' principles, types and tools, detecting and extracting the necessary data from Gaussian® 16 program .log files, preparing the said data to be trained on by a machine learning model, training a non-foundation model for the same problem, fine-tune the foundation model with the prepared data and lastly comparing the results of both models. Results indicate that the XGBoost Regressor model, using internal coordinates, demonstrated higher accuracy with a mean absolute error (MAE) of 0,0026 Hartree. In contrast, the ANI-2x foundation model, trained on Cartesian coordinates, showed a significantly larger MAE of 0,463 Hartree. This suggests that for complex quantum chemistry tasks, foundation models require more extensive data sets and appropriate coordinate transformations to fully realize their potential.","abstract_html":"Understanding and exploring the chemical environment, particularly the dynamics of large molecular structures, is crucial in molecular chemistry. However, the computational resources required for solving such complex systems often necessitate the use of approximate calculation methods, which can compromise result accuracy. To address it, this research explores the application of foundation machine learning models to analyze quantum chemistry data, specifically focusing on predicting molecular energy states accurately without sacrificing computational efficiency. The main problem addressed by this research is the accurate determination of a molecule's ground state energy from its Cartesian coordinates, a task complicated by the need for substantial computational power and precision. The primary objective is to assess the potential application of foundation machine learning models for quantum chemistry problems. The main tasks that are involved in achieving the said objective are understanding the fonudation models' principles, types and tools, detecting and extracting the necessary data from Gaussian® 16 program .log files, preparing the said data to be trained on by a machine learning model, training a non-foundation model for the same problem, fine-tune the foundation model with the prepared data and lastly comparing the results of both models. Results indicate that the XGBoost Regressor model, using internal coordinates, demonstrated higher accuracy with a mean absolute error (MAE) of 0,0026 Hartree. In contrast, the ANI-2x foundation model, trained on Cartesian coordinates, showed a significantly larger MAE of 0,463 Hartree. This suggests that for complex quantum chemistry tasks, foundation models require more extensive data sets and appropriate coordinate transformations to fully realize their potential.","abstract_has_math":false,"creators":["Daranda, Kasparas,"],"institution":"Institutional Repository of Vilnius University","degree_name":null,"degree_level":null,"degree_discipline":null,"degree_department":null,"school":null,"contributors":[],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2024,"date_issued":"2024","date_published":"2024","updated_at":"2026-07-24T05:55:44Z","subjects":[],"languages":["lit"],"rights":["info:eu-repo/semantics/openAccess"],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"https://repository.vu.lt/VU:ELABAETD210578055&prefLang=en_US","outbound_label":"Repository record","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:creator","label":"Author","values":["Daranda, Kasparas,"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2024"]},{"key":"dc:publisher","label":"Institution","values":["Institutional Repository of Vilnius University"]},{"key":"dc:relation","label":"Dc Relation","values":["https://epublications.vu.lt/object/elaba:210578055/210578055.pdf"]},{"key":"dc:type","label":"Dc Type","values":["info:eu-repo/semantics/bachelorThesis"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language","label":"Dc Language","values":["lit"]},{"key":"dc:rights","label":"Dc Rights","values":["info:eu-repo/semantics/openAccess"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["https://repository.vu.lt/VU:ELABAETD210578055&prefLang=en_US"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["Understanding and exploring the chemical environment, particularly the dynamics of large molecular structures, is crucial in molecular chemistry. However, the computational resources required for solving such complex systems often necessitate the use of approximate calculation methods, which can compromise result accuracy. To address it, this research explores the application of foundation machine learning models to analyze quantum chemistry data, specifically focusing on predicting molecular energy states accurately without sacrificing computational efficiency. The main problem addressed by this research is the accurate determination of a molecule's ground state energy from its Cartesian coordinates, a task complicated by the need for substantial computational power and precision. The primary objective is to assess the potential application of foundation machine learning models for quantum chemistry problems. The main tasks that are involved in achieving the said objective are understanding the fonudation models' principles, types and tools, detecting and extracting the necessary data from Gaussian® 16 program .log files, preparing the said data to be trained on by a machine learning model, training a non-foundation model for the same problem, fine-tune the foundation model with the prepared data and lastly comparing the results of both models. Results indicate that the XGBoost Regressor model, using internal coordinates, demonstrated higher accuracy with a mean absolute error (MAE) of 0,0026 Hartree. In contrast, the ANI-2x foundation model, trained on Cartesian coordinates, showed a significantly larger MAE of 0,463 Hartree. This suggests that for complex quantum chemistry tasks, foundation models require more extensive data sets and appropriate coordinate transformations to fully realize their potential."]},{"key":"dc:format","label":"Dc Format","values":["application/pdf"]},{"key":"dc:title","label":"Title","values":["Bazinių mašininio mokymo modelių taikymas kvantinės chemijos duomenų analizei /","Application of machine learning foundation models in quantum-chemical data analysis."]}]}],"canonical_facts":{"dc:creator":["Daranda, Kasparas,"],"dc:date":["2024"],"dc:description":["Understanding and exploring the chemical environment, particularly the dynamics of large molecular structures, is crucial in molecular chemistry. However, the computational resources required for solving such complex systems often necessitate the use of approximate calculation methods, which can compromise result accuracy. To address it, this research explores the application of foundation machine learning models to analyze quantum chemistry data, specifically focusing on predicting molecular energy states accurately without sacrificing computational efficiency. The main problem addressed by this research is the accurate determination of a molecule's ground state energy from its Cartesian coordinates, a task complicated by the need for substantial computational power and precision. The primary objective is to assess the potential application of foundation machine learning models for quantum chemistry problems. The main tasks that are involved in achieving the said objective are understanding the fonudation models' principles, types and tools, detecting and extracting the necessary data from Gaussian® 16 program .log files, preparing the said data to be trained on by a machine learning model, training a non-foundation model for the same problem, fine-tune the foundation model with the prepared data and lastly comparing the results of both models. Results indicate that the XGBoost Regressor model, using internal coordinates, demonstrated higher accuracy with a mean absolute error (MAE) of 0,0026 Hartree. In contrast, the ANI-2x foundation model, trained on Cartesian coordinates, showed a significantly larger MAE of 0,463 Hartree. This suggests that for complex quantum chemistry tasks, foundation models require more extensive data sets and appropriate coordinate transformations to fully realize their potential."],"dc:format":["application/pdf"],"dc:identifier":["https://repository.vu.lt/VU:ELABAETD210578055&prefLang=en_US"],"dc:language":["lit"],"dc:publisher":["Institutional Repository of Vilnius University"],"dc:relation":["https://epublications.vu.lt/object/elaba:210578055/210578055.pdf"],"dc:rights":["info:eu-repo/semantics/openAccess"],"dc:title":["Bazinių mašininio mokymo modelių taikymas kvantinės chemijos duomenų analizei /","Application of machine learning foundation models in quantum-chemical data analysis."],"dc:type":["info:eu-repo/semantics/bachelorThesis"]},"updated_at":"2026-07-24T05:55:44Z"}