Physical data embedding for memory efficient AI
Uloženo v:
| Název: | Physical data embedding for memory efficient AI |
|---|---|
| Autoři: | Callen MacPhee, Yiming Zhou, Bahram Jalali |
| Zdroj: | Machine Learning: Science and Technology, Vol 6, Iss 4, p 045018 (2025) |
| Informace o vydavateli: | IOP Publishing |
| Rok vydání: | 2025 |
| Sbírka: | Directory of Open Access Journals: DOAJ Articles |
| Témata: | physics-AI symbiosis, interpretable AI, physics-inspired algorithms, physics-based neural networks, memory-efficient AI, Computer engineering. Computer hardware, TK7885-7895, Electronic computers. Computer science, QA75.5-76.95 |
| Popis: | Deep neural networks have achieved exceptional performance across various fields by learning complex, nonlinear mappings from large-scale datasets. However, they face challenges such as high memory requirements and limited interpretability. This paper introduces an approach where master equations of physics are converted into multilayered networks that are trained via backpropagation. The resulting general-purpose model effectively encodes data in the properties of the underlying physical system. In contrast to existing methods wherein a trained neural network is used as a computationally efficient alternative for solving physical equations, our approach directly treats physics equations as trainable models. Rather than approximating physics with a neural network or augmenting a network with physics-inspired constraints, this framework makes the equation itself the architecture. We demonstrate this physical embedding concept with the nonlinear Schrödinger equation, which acts as trainable architecture for learning complex patterns including nonlinear mappings and memory effects from data. The network embeds data representation in orders of magnitude fewer parameters than conventional neural networks when tested on time series data. Notably, the trained ‘Nonlinear Schrödinger Network’ is interpretable, with all parameters having physical meanings. Curiously, this approach also provides a blueprint for implementing such AI computations in physical analog systems, offering a direct path toward low-latency and energy-efficient hardware realizations. The proposed method is also extended to the Gross-Pitaevskii equation, demonstrating the broad applicability of the framework to other master equations of physics. Among our results, an ablation study quantifies the relative importance of physical terms such as dispersion, nonlinearity, and potential energy for classification accuracy. We also outline the limitations and benefits of this approach as it relates to universality and generalizability. Overall, this work aims ... |
| Druh dokumentu: | article in journal/newspaper |
| Jazyk: | English |
| Relation: | https://doi.org/10.1088/2632-2153/ae0f37; https://doaj.org/toc/2632-2153; https://doaj.org/article/b97d56612e9342afadd949cf144179ee |
| DOI: | 10.1088/2632-2153/ae0f37 |
| Dostupnost: | https://doi.org/10.1088/2632-2153/ae0f37 https://doaj.org/article/b97d56612e9342afadd949cf144179ee |
| Přístupové číslo: | edsbas.238D9702 |
| Databáze: | BASE |
Nájsť tento článok vo Web of Science | FullText | Text: Availability: 0 CustomLinks: – Url: https://doi.org/10.1088/2632-2153/ae0f37# Name: EDS - BASE (s4221598) Category: fullText Text: View record from BASE – Url: https://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=EBSCO&SrcAuth=EBSCO&DestApp=WOS&ServiceName=TransferToWoS&DestLinkType=GeneralSearchSummary&Func=Links&author=MacPhee%20C Name: ISI Category: fullText Text: Nájsť tento článok vo Web of Science Icon: https://imagesrvr.epnet.com/ls/20docs.gif MouseOverText: Nájsť tento článok vo Web of Science |
|---|---|
| Header | DbId: edsbas DbLabel: BASE An: edsbas.238D9702 RelevancyScore: 997 AccessLevel: 3 PubType: Academic Journal PubTypeId: academicJournal PreciseRelevancyScore: 996.707214355469 |
| IllustrationInfo | |
| Items | – Name: Title Label: Title Group: Ti Data: Physical data embedding for memory efficient AI – Name: Author Label: Authors Group: Au Data: <searchLink fieldCode="AR" term="%22Callen+MacPhee%22">Callen MacPhee</searchLink><br /><searchLink fieldCode="AR" term="%22Yiming+Zhou%22">Yiming Zhou</searchLink><br /><searchLink fieldCode="AR" term="%22Bahram+Jalali%22">Bahram Jalali</searchLink> – Name: TitleSource Label: Source Group: Src Data: Machine Learning: Science and Technology, Vol 6, Iss 4, p 045018 (2025) – Name: Publisher Label: Publisher Information Group: PubInfo Data: IOP Publishing – Name: DatePubCY Label: Publication Year Group: Date Data: 2025 – Name: Subset Label: Collection Group: HoldingsInfo Data: Directory of Open Access Journals: DOAJ Articles – Name: Subject Label: Subject Terms Group: Su Data: <searchLink fieldCode="DE" term="%22physics-AI+symbiosis%22">physics-AI symbiosis</searchLink><br /><searchLink fieldCode="DE" term="%22interpretable+AI%22">interpretable AI</searchLink><br /><searchLink fieldCode="DE" term="%22physics-inspired+algorithms%22">physics-inspired algorithms</searchLink><br /><searchLink fieldCode="DE" term="%22physics-based+neural+networks%22">physics-based neural networks</searchLink><br /><searchLink fieldCode="DE" term="%22memory-efficient+AI%22">memory-efficient AI</searchLink><br /><searchLink fieldCode="DE" term="%22Computer+engineering%2E+Computer+hardware%22">Computer engineering. Computer hardware</searchLink><br /><searchLink fieldCode="DE" term="%22TK7885-7895%22">TK7885-7895</searchLink><br /><searchLink fieldCode="DE" term="%22Electronic+computers%2E+Computer+science%22">Electronic computers. Computer science</searchLink><br /><searchLink fieldCode="DE" term="%22QA75%2E5-76%2E95%22">QA75.5-76.95</searchLink> – Name: Abstract Label: Description Group: Ab Data: Deep neural networks have achieved exceptional performance across various fields by learning complex, nonlinear mappings from large-scale datasets. However, they face challenges such as high memory requirements and limited interpretability. This paper introduces an approach where master equations of physics are converted into multilayered networks that are trained via backpropagation. The resulting general-purpose model effectively encodes data in the properties of the underlying physical system. In contrast to existing methods wherein a trained neural network is used as a computationally efficient alternative for solving physical equations, our approach directly treats physics equations as trainable models. Rather than approximating physics with a neural network or augmenting a network with physics-inspired constraints, this framework makes the equation itself the architecture. We demonstrate this physical embedding concept with the nonlinear Schrödinger equation, which acts as trainable architecture for learning complex patterns including nonlinear mappings and memory effects from data. The network embeds data representation in orders of magnitude fewer parameters than conventional neural networks when tested on time series data. Notably, the trained ‘Nonlinear Schrödinger Network’ is interpretable, with all parameters having physical meanings. Curiously, this approach also provides a blueprint for implementing such AI computations in physical analog systems, offering a direct path toward low-latency and energy-efficient hardware realizations. The proposed method is also extended to the Gross-Pitaevskii equation, demonstrating the broad applicability of the framework to other master equations of physics. Among our results, an ablation study quantifies the relative importance of physical terms such as dispersion, nonlinearity, and potential energy for classification accuracy. We also outline the limitations and benefits of this approach as it relates to universality and generalizability. Overall, this work aims ... – Name: TypeDocument Label: Document Type Group: TypDoc Data: article in journal/newspaper – Name: Language Label: Language Group: Lang Data: English – Name: NoteTitleSource Label: Relation Group: SrcInfo Data: https://doi.org/10.1088/2632-2153/ae0f37; https://doaj.org/toc/2632-2153; https://doaj.org/article/b97d56612e9342afadd949cf144179ee – Name: DOI Label: DOI Group: ID Data: 10.1088/2632-2153/ae0f37 – Name: URL Label: Availability Group: URL Data: https://doi.org/10.1088/2632-2153/ae0f37<br />https://doaj.org/article/b97d56612e9342afadd949cf144179ee – Name: AN Label: Accession Number Group: ID Data: edsbas.238D9702 |
| PLink | https://erproxy.cvtisr.sk/sfx/access?url=https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=edsbas&AN=edsbas.238D9702 |
| RecordInfo | BibRecord: BibEntity: Identifiers: – Type: doi Value: 10.1088/2632-2153/ae0f37 Languages: – Text: English Subjects: – SubjectFull: physics-AI symbiosis Type: general – SubjectFull: interpretable AI Type: general – SubjectFull: physics-inspired algorithms Type: general – SubjectFull: physics-based neural networks Type: general – SubjectFull: memory-efficient AI Type: general – SubjectFull: Computer engineering. Computer hardware Type: general – SubjectFull: TK7885-7895 Type: general – SubjectFull: Electronic computers. Computer science Type: general – SubjectFull: QA75.5-76.95 Type: general Titles: – TitleFull: Physical data embedding for memory efficient AI Type: main BibRelationships: HasContributorRelationships: – PersonEntity: Name: NameFull: Callen MacPhee – PersonEntity: Name: NameFull: Yiming Zhou – PersonEntity: Name: NameFull: Bahram Jalali IsPartOfRelationships: – BibEntity: Dates: – D: 01 M: 01 Type: published Y: 2025 Identifiers: – Type: issn-locals Value: edsbas – Type: issn-locals Value: edsbas.oa Titles: – TitleFull: Machine Learning: Science and Technology, Vol 6, Iss 4, p 045018 (2025 Type: main |
| ResultId | 1 |