Show simple item record

Enhancing Physical Modeling with Interpretable Physics-Aware Machine Learning

dc.contributor.authorJacobsen, Christian
dc.date.accessioned2024-05-22T17:27:54Z
dc.date.available2024-05-22T17:27:54Z
dc.date.issued2024
dc.date.submitted2024
dc.identifier.urihttps://hdl.handle.net/2027.42/193425
dc.description.abstractThe burgeoning intersection of machine learning (ML) with physics has catalyzed a transformative approach to physical modeling marked by an enhanced capacity for innovation and discovery. Traditional applications of ML in physics often grapple with a critical challenge: predictions frequently lack transparency and interpretability, ultimately compromising generalizability and hindering the diagnosis of limitations. This opacity reduces the ability to develop generalized physical models and extract deeper insights into the underlying physical processes. This work aims to harness the capabilities of ML to learn from data while enhancing the interpretability of its models by creating physics-aware models. In doing so, it seeks to facilitate deeper analyses and enable more profound insights into physical phenomena, developing models which improve accuracy and generalizability over baseline methods. Through a series of studies, this work delves into the development and application of physically interpretable models in the realm of physics-aware machine learning. Each of the works, while distinct in their focus and application, collectively contributes to the advancement of interpretable physics-informed modeling with machine learning, addressing both theoretical and practical aspects. Together, these studies develop a narrative that underscores the importance of interpretability in machine learning for enhancing the generalizability of models. Additionally, it focuses on the development of physics-aware machine learning models to achieve this objective. Each piece of research, distinct in its application, contributes to a unified theme of enhancing the interpretability, generalizability, and practical utility of machine learning in the realm of physics. This body of work advances our understanding of how machine learning can be effectively applied to physics problems and offers new perspectives and methodologies that can aid in many aspects of scientific exploration and discovery.
dc.language.isoen_US
dc.subjectComputational science
dc.subjectMachine Learning
dc.subjectPhysics-aware
dc.subjectPhysics-informed
dc.titleEnhancing Physical Modeling with Interpretable Physics-Aware Machine Learning
dc.typeThesis
dc.description.thesisdegreenamePhD
dc.description.thesisdegreedisciplineAerospace Engineering
dc.description.thesisdegreegrantorUniversity of Michigan, Horace H. Rackham School of Graduate Studies
dc.contributor.committeememberDuraisamy, Karthik
dc.contributor.committeememberHuan, Xun
dc.contributor.committeememberFidkowski, Krzysztof J
dc.contributor.committeememberGorodetsky, Alex Arkady
dc.subject.hlbsecondlevelAerospace Engineering
dc.subject.hlbsecondlevelComputer Science
dc.subject.hlbsecondlevelPhysics
dc.subject.hlbtoplevelEngineering
dc.subject.hlbtoplevelScience
dc.contributor.affiliationumcampusAnn Arbor
dc.description.bitstreamurlhttp://deepblue.lib.umich.edu/bitstream/2027.42/193425/1/csjacobs_1.pdf
dc.identifier.doihttps://dx.doi.org/10.7302/23070
dc.identifier.orcid0000-0002-1394-5893
dc.identifier.name-orcidJacobsen, Christian; 0000-0002-1394-5893en_US
dc.working.doi10.7302/23070en
dc.owningcollnameDissertations and Theses (Ph.D. and Master's)


Files in this item

Show simple item record

Remediation of Harmful Language

The University of Michigan Library aims to describe library materials in a way that respects the people and communities who create, use, and are represented in our collections. Report harmful or offensive language in catalog records, finding aids, or elsewhere in our collections anonymously through our metadata feedback form. More information at Remediation of Harmful Language.

Accessibility

If you are unable to use this file in its current format, please select the Contact Us link and we can modify it to make it more accessible to you.