{"id":{"repo_id":"iastate","oai_identifier":"oai:dr.lib.iastate.edu:20.500.12876/KrZJOYXr"},"canonical_url":"https://search.dev.ndltd.org/etd/iastate/oai:dr.lib.iastate.edu:20.500.12876/KrZJOYXr","repository":{"repo_id":"iastate","name":"Iowa State University","base_url":"https://dr.lib.iastate.edu/server/oai/request"},"display":{"title":"AI for materials design: Generative AI with multi-fidelity strategies","abstract":"The design of new molecules and materials is often hindered by the vastness of chemical and microstructural design spaces and the high cost of obtaining high-fidelity property labels through quantum or physics-based simulations. This dissertation introduces a unified framework that combines generative artificial intelligence (AI), hierarchical transfer learning, multi-fidelity modeling, and graph-driven voxel-based analysis to address four critical challenges in materials discovery: the need for syntactically valid and interpretable generative models, the data inefficiency of high-fidelity property prediction, the unreliability of models under distribution shift, and the lack of scalable tools for characterizing dynamic structural domains. First, we present MolGen-Transformer, a transformer-based molecular language model trained on a dataset of 198 million molecules using the SELFIES representation. It achieves perfect reconstruction accuracy, generates chemically diverse and valid molecules, and offers a compact and interpretable latent space suitable for downstream design tasks. Second, we build on this well-trained model by developing a hierarchical property prediction framework that fuses MolGen-Transformer embeddings with both low- and high-fidelity labels. This multi-fidelity approach reduces the dependence on expensive density functional theory (DFT) data by up to fourfold. It also integrates uncertainty quantification via ensemble modeling to support reliable, property-driven molecular design through latent space path search. Third, we generalize the multi-fidelity paradigm beyond molecules by applying it to microstructure–property prediction for organic photovoltaics. We demonstrate that a combination of learned microstructure embeddings and limited high-fidelity simulations enables accurate prediction of device-level performance with data efficiency. Fourth, we introduce MDVoxelizer, a modular framework that integrates graph-based structural filtering with voxel-based spatial mapping to quantify local crystallinity in molecular dynamics simulations. This enables interpretable, time-resolved, and machine learning–compatible representations of complex, evolving molecular systems. Collectively, these contributions establish a robust, generalizable approach to AI-guided design. By tightly integrating generative modeling, fidelity-aware learning, structural characterization, and uncertainty estimation, this work enables scalable exploration and optimization in chemical and materials spaces while mitigating computational cost and predictive risk.","abstract_html":"The design of new molecules and materials is often hindered by the vastness of chemical and microstructural design spaces and the high cost of obtaining high-fidelity property labels through quantum or physics-based simulations. This dissertation introduces a unified framework that combines generative artificial intelligence (AI), hierarchical transfer learning, multi-fidelity modeling, and graph-driven voxel-based analysis to address four critical challenges in materials discovery: the need for syntactically valid and interpretable generative models, the data inefficiency of high-fidelity property prediction, the unreliability of models under distribution shift, and the lack of scalable tools for characterizing dynamic structural domains. First, we present MolGen-Transformer, a transformer-based molecular language model trained on a dataset of 198 million molecules using the SELFIES representation. It achieves perfect reconstruction accuracy, generates chemically diverse and valid molecules, and offers a compact and interpretable latent space suitable for downstream design tasks. Second, we build on this well-trained model by developing a hierarchical property prediction framework that fuses MolGen-Transformer embeddings with both low- and high-fidelity labels. This multi-fidelity approach reduces the dependence on expensive density functional theory (DFT) data by up to fourfold. It also integrates uncertainty quantification via ensemble modeling to support reliable, property-driven molecular design through latent space path search. Third, we generalize the multi-fidelity paradigm beyond molecules by applying it to microstructure–property prediction for organic photovoltaics. We demonstrate that a combination of learned microstructure embeddings and limited high-fidelity simulations enables accurate prediction of device-level performance with data efficiency. Fourth, we introduce MDVoxelizer, a modular framework that integrates graph-based structural filtering with voxel-based spatial mapping to quantify local crystallinity in molecular dynamics simulations. This enables interpretable, time-resolved, and machine learning–compatible representations of complex, evolving molecular systems. Collectively, these contributions establish a robust, generalizable approach to AI-guided design. By tightly integrating generative modeling, fidelity-aware learning, structural characterization, and uncertainty estimation, this work enables scalable exploration and optimization in chemical and materials spaces while mitigating computational cost and predictive risk.","abstract_has_math":false,"creators":["Yang, Chih Hsuan"],"institution":"Iowa State University","degree_name":"Doctor of Philosophy","degree_level":"dissertation","degree_discipline":"Artificial intelligence","degree_department":"Department of Mechanical Engineering","school":null,"contributors":[],"advisors":["Ganapathysubramanian, Baskar","Sarkar, Soumik","Krishnamurthy, Adarsh","Liu, Hailiang","Li, Qi"],"committee_chairs":[],"committee_members":[],"year":2025,"date_issued":"2025-12","date_published":"2025-12","updated_at":"2026-07-24T02:37:49Z","subjects":[],"languages":["en"],"rights":[],"rights_urls":[],"identifier_entries":[{"key":"dc:identifier.doi","label":"DOI","values":["https://doi.org/10.31274/td-20260223-95"],"render_values":[{"text":"https://doi.org/10.31274/td-20260223-95","href":"https://doi.org/10.31274/td-20260223-95","code":true}]}]},"links":{"outbound_url":"https://dr.lib.iastate.edu/handle/20.500.12876/KrZJOYXr","outbound_label":"Repository record","outbound_source":"dc:identifier.uri"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor.advisor","label":"Advisor","values":["Ganapathysubramanian, Baskar","Sarkar, Soumik","Krishnamurthy, Adarsh","Liu, Hailiang","Li, Qi"]},{"key":"dc:contributor.department","label":"Department","values":["Department of Mechanical Engineering"]},{"key":"dc:creator","label":"Author","values":["Yang, Chih Hsuan"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.accessioned","label":"Dc Date Accessioned","values":["2026-01-30T17:02:38Z"]},{"key":"dc:date.available","label":"Dc Date Available","values":["2026-01-30T17:02:38Z"]},{"key":"dc:date.issued","label":"Date","values":["2025-12"]},{"key":"dc:type","label":"Dc Type","values":["dissertation"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["Artificial intelligence"]},{"key":"thesis:degree_level","label":"Degree Level","values":["dissertation"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Doctor of Philosophy"]},{"key":"thesis:institution_name","label":"Thesis Institution Name","values":["Iowa State University"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language.iso","label":"Language (ISO)","values":["en"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier.doi","label":"DOI","values":["https://doi.org/10.31274/td-20260223-95"]},{"key":"dc:identifier.uri","label":"Identifier URI","values":["https://dr.lib.iastate.edu/handle/20.500.12876/KrZJOYXr"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["The design of new molecules and materials is often hindered by the vastness of chemical and microstructural design spaces and the high cost of obtaining high-fidelity property labels through quantum or physics-based simulations. This dissertation introduces a unified framework that combines generative artificial intelligence (AI), hierarchical transfer learning, multi-fidelity modeling, and graph-driven voxel-based analysis to address four critical challenges in materials discovery: the need for syntactically valid and interpretable generative models, the data inefficiency of high-fidelity property prediction, the unreliability of models under distribution shift, and the lack of scalable tools for characterizing dynamic structural domains. First, we present MolGen-Transformer, a transformer-based molecular language model trained on a dataset of 198 million molecules using the SELFIES representation. It achieves perfect reconstruction accuracy, generates chemically diverse and valid molecules, and offers a compact and interpretable latent space suitable for downstream design tasks. Second, we build on this well-trained model by developing a hierarchical property prediction framework that fuses MolGen-Transformer embeddings with both low- and high-fidelity labels. This multi-fidelity approach reduces the dependence on expensive density functional theory (DFT) data by up to fourfold. It also integrates uncertainty quantification via ensemble modeling to support reliable, property-driven molecular design through latent space path search. Third, we generalize the multi-fidelity paradigm beyond molecules by applying it to microstructure–property prediction for organic photovoltaics. We demonstrate that a combination of learned microstructure embeddings and limited high-fidelity simulations enables accurate prediction of device-level performance with data efficiency. Fourth, we introduce MDVoxelizer, a modular framework that integrates graph-based structural filtering with voxel-based spatial mapping to quantify local crystallinity in molecular dynamics simulations. This enables interpretable, time-resolved, and machine learning–compatible representations of complex, evolving molecular systems. Collectively, these contributions establish a robust, generalizable approach to AI-guided design. By tightly integrating generative modeling, fidelity-aware learning, structural characterization, and uncertainty estimation, this work enables scalable exploration and optimization in chemical and materials spaces while mitigating computational cost and predictive risk."]},{"key":"dc:format.mimetype","label":"Dc Format Mimetype","values":["PDF"]},{"key":"dc:title","label":"Title","values":["AI for materials design: Generative AI with multi-fidelity strategies"]}]}],"canonical_facts":{"dc:contributor.advisor":["Ganapathysubramanian, Baskar","Sarkar, Soumik","Krishnamurthy, Adarsh","Liu, Hailiang","Li, Qi"],"dc:contributor.department":["Department of Mechanical Engineering"],"dc:creator":["Yang, Chih Hsuan"],"dc:date.accessioned":["2026-01-30T17:02:38Z"],"dc:date.available":["2026-01-30T17:02:38Z"],"dc:date.issued":["2025-12"],"dc:description.abstract":["The design of new molecules and materials is often hindered by the vastness of chemical and microstructural design spaces and the high cost of obtaining high-fidelity property labels through quantum or physics-based simulations. This dissertation introduces a unified framework that combines generative artificial intelligence (AI), hierarchical transfer learning, multi-fidelity modeling, and graph-driven voxel-based analysis to address four critical challenges in materials discovery: the need for syntactically valid and interpretable generative models, the data inefficiency of high-fidelity property prediction, the unreliability of models under distribution shift, and the lack of scalable tools for characterizing dynamic structural domains. First, we present MolGen-Transformer, a transformer-based molecular language model trained on a dataset of 198 million molecules using the SELFIES representation. It achieves perfect reconstruction accuracy, generates chemically diverse and valid molecules, and offers a compact and interpretable latent space suitable for downstream design tasks. Second, we build on this well-trained model by developing a hierarchical property prediction framework that fuses MolGen-Transformer embeddings with both low- and high-fidelity labels. This multi-fidelity approach reduces the dependence on expensive density functional theory (DFT) data by up to fourfold. It also integrates uncertainty quantification via ensemble modeling to support reliable, property-driven molecular design through latent space path search. Third, we generalize the multi-fidelity paradigm beyond molecules by applying it to microstructure–property prediction for organic photovoltaics. We demonstrate that a combination of learned microstructure embeddings and limited high-fidelity simulations enables accurate prediction of device-level performance with data efficiency. Fourth, we introduce MDVoxelizer, a modular framework that integrates graph-based structural filtering with voxel-based spatial mapping to quantify local crystallinity in molecular dynamics simulations. This enables interpretable, time-resolved, and machine learning–compatible representations of complex, evolving molecular systems. Collectively, these contributions establish a robust, generalizable approach to AI-guided design. By tightly integrating generative modeling, fidelity-aware learning, structural characterization, and uncertainty estimation, this work enables scalable exploration and optimization in chemical and materials spaces while mitigating computational cost and predictive risk."],"dc:format.mimetype":["PDF"],"dc:identifier.doi":["https://doi.org/10.31274/td-20260223-95"],"dc:identifier.uri":["https://dr.lib.iastate.edu/handle/20.500.12876/KrZJOYXr"],"dc:language.iso":["en"],"dc:title":["AI for materials design: Generative AI with multi-fidelity strategies"],"dc:type":["dissertation"],"thesis:degree_discipline":["Artificial intelligence"],"thesis:degree_level":["dissertation"],"thesis:degree_name":["Doctor of Philosophy"],"thesis:institution_name":["Iowa State University"]},"updated_at":"2026-07-24T02:37:49Z"}