000625360 001__ 625360
000625360 005__ 20250410223119.0
000625360 037__ $$aPUBDB-2025-01117
000625360 041__ $$aEnglish
000625360 1001_ $$0P:(DE-H253)PIP1102179$$aFernandez Corral, Alvaro$$b0$$eCorresponding author$$udesy
000625360 1112_ $$aInternal Conference on Scientific Computing and Machine Learning$$cKyoto$$d2025-03-03 - 2025-03-07$$gSCML$$wJapan
000625360 245__ $$aThe coordinate is right - Augmenting basis sets via normalizing flows
000625360 260__ $$c2025
000625360 3367_ $$033$$2EndNote$$aConference Paper
000625360 3367_ $$2DataCite$$aOther
000625360 3367_ $$2BibTeX$$aINPROCEEDINGS
000625360 3367_ $$2DRIVER$$aconferenceObject
000625360 3367_ $$2ORCID$$aLECTURE_SPEECH
000625360 3367_ $$0PUB:(DE-HGF)6$$2PUB:(DE-HGF)$$aConference Presentation$$bconf$$mconf$$s1744289096_3645141$$xInvited
000625360 520__ $$aApproximating functions using a basis set of the functions’ space guarantees convergence of the approximation. However, high-accuracy calculations require memory costs that scale exponentially with the dimensionality, formally known as the curse of dimensionality. In this talk, I will introduce augmented basis sets, generated by pushing forward standard basis sets through normalizing flows, i.e., invertible neural networks, which is equivalent to a transformation of the basis’ coordinates. Multidimensional basis sets are often built from direct-products of univariate functions. These constructions struggle to capture complex structures involving different coordinates. Normalizing-flows coordinates reduce the coupling between dimensions, mitigating the curse of dimensionality.I will demonstrate the efficacy of augmented basis sets to approximate eigenpairs of the vibrational Schrödinger equation. Unlike standard neural-network-based methods that directly model the eigenfunctions, our method preserves the basis properties, ensuring robustness in the approximation of many highly excited states. Additionally, optimal normalizing-flows coordinates encode physical information of the molecular motion, which allows for the interpretability of the method, and enables transferability to different basis set truncations sizes and even to structurally similar molecular systems.
000625360 536__ $$0G:(DE-HGF)POF4-631$$a631 - Matter – Dynamics, Mechanisms and Control (POF4-631)$$cPOF4-631$$fPOF IV$$x0
000625360 536__ $$0G:(DE-HGF)2019_IVF-HIDSS-0002$$aHIDSS-0002 - DASHH: Data Science in Hamburg - Helmholtz Graduate School for the Structure of Matter (2019_IVF-HIDSS-0002)$$c2019_IVF-HIDSS-0002$$x1
000625360 693__ $$0EXP:(DE-MLZ)NOSPEC-20140101$$5EXP:(DE-MLZ)NOSPEC-20140101$$eNo specific instrument$$x0
000625360 8564_ $$uhttps://scml.jp/scml2025_program.pdf
000625360 8564_ $$uhttps://bib-pubdb1.desy.de/record/625360/files/scml2025_program.pdf$$yRestricted
000625360 8564_ $$uhttps://bib-pubdb1.desy.de/record/625360/files/scml2025_program.pdf?subformat=pdfa$$xpdfa$$yRestricted
000625360 909CO $$ooai:bib-pubdb1.desy.de:625360$$pVDB
000625360 9101_ $$0I:(DE-588b)2008985-5$$6P:(DE-H253)PIP1102179$$aDeutsches Elektronen-Synchrotron$$b0$$kDESY
000625360 9101_ $$0I:(DE-H253)_CFEL-20120731$$6P:(DE-H253)PIP1102179$$aCentre for Free-Electron Laser Science$$b0$$kCFEL
000625360 9131_ $$0G:(DE-HGF)POF4-631$$1G:(DE-HGF)POF4-630$$2G:(DE-HGF)POF4-600$$3G:(DE-HGF)POF4$$4G:(DE-HGF)POF$$aDE-HGF$$bForschungsbereich Materie$$lVon Materie zu Materialien und Leben$$vMatter – Dynamics, Mechanisms and Control$$x0
000625360 9141_ $$y2025
000625360 9201_ $$0I:(DE-H253)FS-CFEL-CMI-20220405$$kFS-CFEL-CMI$$lCFEL-CMI$$x0
000625360 980__ $$aconf
000625360 980__ $$aVDB
000625360 980__ $$aI:(DE-H253)FS-CFEL-CMI-20220405
000625360 980__ $$aUNRESTRICTED