检索结果-内蒙古大学图书馆

NOMAD: nonlinear manifold decoders for operator learning 22

学校读者我要写书评

暂无评论

NOMAD: nonlinear manifold decoders for operator learning

Proceedings of the 36th International Conference on Neural Information Processing Systems

作者： Jacob H. Seidman Georgios Kissas Paris Perdikaris George J. Pappas Graduate Program in Applied Mathematics and Computational Science University of Pennsylvania Department of Mechanical Engineering and Applied Mechanics University of Pennsylvania Department of Electrical and Systems Engineering University of Pennsylvania

ISBN: (纸本)9781713871088

Supervised learning in function spaces is an emerging area of machine learning research with applications to the prediction of complex physical systems such as fluid flows, solid mechanics, and climate modeling. By directly learning maps (operators) between infinite dimensional function spaces, these models are able to learn discretization invariant representations of target functions. A common approach is to represent such target functions as linear combinations of basis elements learned from data. However, there are simple scenarios where, even though the target functions form a low dimensional submanifold, a very large number of basis elements is needed for an accurate linear representation. Here we present NOMAD, a novel operator learning framework with a nonlinear decoder map capable of learning finite dimensional representations of nonlinear submanifolds in function spaces. We show this method is able to accurately learn low dimensional representations of solution manifolds to partial differential equations while outperforming linear models of larger size. Additionally, we compare to state-of-the-art operator learning methods on a complex fluid dynamics benchmark and achieve competitive performance with a significantly smaller model size and training cost.

关键词：

Inferring System and Optimal Control Parameters of Closed-Loop Systems from Partial Observations

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Geadah, Victor Arbelaiz, Juncal Ritz, Harrison Daw, Nathaniel D. Cohen, Jonathan D. Pillow, Jonathan W. Program in Applied and Computational Mathematics Princeton University PrincetonNJ United States Department of Mechanical and Aerospace Engineering Princeton University PrincetonNJ United States Princeton Neuroscience Institute Princeton University PrincetonNJ United States

We consider the joint problem of system identification and inverse optimal control for discrete-time stochastic Linear Quadratic Regulators. We analyze finite and infinite time horizons in a partially observed setting, where the state is observed noisily. To recover closed-loop system parameters, we develop inference methods based on probabilistic state-space model (SSM) techniques. First, we show that the system parameters exhibit non-identifiability in the infinite-horizon from closed-loop measurements, and we provide exact and numerical methods to disentangle the parameters. Second, to improve parameter identifiability, we show that we can further enhance recovery by either (1) incorporating additional partial measurements of the control signals or (2) moving to the finite-horizon setting. We further illustrate the performance of our methodology through numerical examples. © 2025, CC BY.

关键词： Stochastic systems

Viscosity Solutions of the Eikonal Equation on the Wasserstein Space∗

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Soner, H. Mete Yan, Qinxin Department of Operations Research and Financial Engineering Princeton University PrincetonNJ08540 United States Program in Applied and Computational Mathematics Princeton University PrincetonNJ08540 United States

Dynamic programming equations for mean field control problems with a separable structure are Eikonal type equations on the Wasserstein space. Standard differentiation using linear derivatives yield a direct extension of the classical viscosity theory. We use Fourier representation of the Sobolev norms on the space of measures, together with the standard techniques from the finite dimensional theory to prove a comparison result among semicontinuous sub and super solutions, obtaining a unique characterization of the value *** Codes 35D40, 35Q89, 49L12, 49L25, 60G99 © 2023, CC BY-NC-ND.

关键词： Viscosity

Entropic Optimal Transport Eigenmaps for Nonlinear Alignment and Joint Embedding of High-Dimensional Datasets

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Landa, Boris Kluger, Yuval Ma, Rong Department of Electrical Engineering Yale University United States Program in Applied Mathematics Yale University United States Department of Biostatistics Harvard University United States Interdepartmental Program in Computational Biology and Bioinformatics Yale University United States Department of Pathology Yale University School of Medicine United States

Embedding high-dimensional data into a low-dimensional space is an indispensable component of data analysis. In numerous applications, it is necessary to align and jointly embed multiple datasets from different studies or experimental conditions. Such datasets may share underlying structures of interest but exhibit individual distortions, resulting in misaligned embeddings using traditional techniques. In this work, we propose Entropic Optimal Transport (EOT) eigenmaps, a principled approach for aligning and jointly embedding a pair of datasets with theoretical guarantees. Our approach leverages the leading singular vectors of the EOT plan matrix between two datasets to extract their shared underlying structure and align the datasets accordingly in a common embedding space. We interpret our approach as an inter-data variant of the classical Laplacian eigenmaps and diffusion maps embeddings, showing that it enjoys many favorable analogous properties. We then analyze a data-generative model where two observed high-dimensional datasets share latent variables on a common low-dimensional manifold, but each dataset is subject to data-specific translation, scaling, nuisance structures, and noise. We show that in a high-dimensional asymptotic regime, the EOT plan recovers the shared manifold structure by approximating a kernel function evaluated at the locations of the latent variables. Subsequently, we provide a geometric interpretation of our embedding by relating it to the eigenfunctions of population-level operators encoding the density and geometry of the shared manifold. Finally, we showcase the performance of our approach for data integration and embedding through simulations and analyses of real-world biological data, demonstrating its advantages over alternative methods in challenging scenarios. Copyright © 2024, The Authors. All rights reserved.

关键词： Embeddings

Ground state phases of the two-dimension electron gas with a unified variational approach

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Smith, Conor Chen, Yixiao Levy, Ryan Yang, Yubo Morales, Miguel A. Zhang, Shiwei Center for Computational Quantum Physics Flatiron Institute New YorkNY10010 United States Department of Electrical and Computer Engineering University of New Mexico AlbuquerqueNM87131 United States Program in Applied and Computational Mathematics Princeton University PrincetonNJ08544 United States

The two-dimensional electron gas (2DEG) is a fundamental model, which is drawing increasing interest because of recent advances in experimental and theoretical studies of 2D materials. Current understanding of the ground state of the 2DEG relies on quantum Monte Carlo calculations, based on variational comparisons of different ansatze for different phases. We use a single variational ansatz, a general backflow-type wave function using a message-passing neural quantum state architecture, for a unified description across the entire density range. The variational optimization consistently leads to lower ground-state energies than previous best results. Transition into a Wigner crystal (WC) phase occurs automatically at rs = 37±1, a density lower than currently believed. Between the liquid and WC phases, the same ansatz and variational search strongly suggest the existence of intermediate states in a broad range of densities, with enhanced short-range nematic spin correlations. Copyright © 2024, The Authors. All rights reserved.

关键词： Message passing

Viscoelastic fluid flow in a slowly varying planar contraction: the role of finite extensibility on the pressure drop

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Mahapatra, Bimalendu Ruangkriengsin, Tachin Stone, Howard A. Boyko, Evgeniy Faculty of Mechanical Engineering Technion – Israel Institute of Technology Haifa3200003 Israel Program in Applied and Computational Mathematics Princeton University PrincetonNJ08544 United States Department of Mechanical and Aerospace Engineering Princeton University PrincetonNJ08544 United States

We analyze the steady viscoelastic fluid flow in slowly varying contracting channels of arbitrary shape and present a theory based on the lubrication approximation for calculating the flow rate–pressure drop relation at low and high Deborah (De) numbers. Unlike most prior theoretical studies leveraging the Oldroyd-B model, we describe the fluid viscoelasticity using a FENE-CR model and examine how the polymer chains’ finite extensibility impacts the pressure drop. We employ the low-Deborah-number lubrication analysis to provide analytical expressions for the pressure drop up to O(De4). We further consider the ultra-dilute limit and exploit a one-way coupling between the parabolic velocity and elastic stresses to calculate the pressure drop of the FENE-CR fluid for arbitrary values of the Deborah number. Such an approach allows us to elucidate elastic stress contributions governing the pressure drop variations and the effect of finite extensibility for all De. We validate our theoretical predictions with two-dimensional numerical simulations and find excellent agreement. We show that, at low Deborah numbers, the pressure drop of the FENE-CR fluid monotonically decreases with De, similar to the previous results for the Oldroyd-B and FENE-P fluids. However, at high Deborah numbers, in contrast to a linear decrease for the Oldroyd-B fluid, the pressure drop of the FENE-CR fluid exhibits a non-monotonic variation due to finite extensibility, first decreasing and then increasing with de. Nevertheless, even at sufficiently high Deborah numbers, the pressure drop of the FENE-CR fluid in the ultra-dilute and lubrication limits is lower than the corresponding Newtonian pressure drop. © 2024, CC BY.

关键词： Non Newtonian flow

Advancing electrochemical impedance analysis through innovations in the distribution of relaxation times method

学校读者我要写书评

暂无评论

Joule 2024年第7期8卷 1958-1981页

作者： Maradesa, Adeleke Py, Baptiste Huang, Jake Lu, Yang Iurilli, Pietro Mrozinski, Aleksander Law, Ho Mei Wang, Yuhao Wang, Zilong Li, Jingwei Xu, Shengjun Meyer, Quentin Liu, Jiapeng Brivio, Claudio Gavrilyuk, Alexander Kobayashi, Kiyoshi Bertei, Antonio Williams, Nicholas J. Zhao, Chuan Danzer, Michael Zic, Mark Wu, Phillip Yrjänä, Ville Pereverzyev, Sergei Chen, Yuhui Weber, André Kalinin, Sergei V. Schmidt, Jan Philipp Tsur, Yoed Boukamp, Bernard A. Zhang, Qiang Gaberšček, Miran O'Hayre, Ryan Ciucci, Francesco Department of Mechanical and Aerospace Engineering The Hong Kong University of Science and Technology Hong Kong Metallurgical and Materials Engineering Colorado School of Mines Golden80401 United States Beijing Key Laboratory of Green Chemical Reaction Engineering and Technology Department of Chemical Engineering Tsinghua University Beijing China Green Energy Storage Trento38123 Italy Department of Manufacturing and Production Engineering Faculty of Mechanical Engineering and Ship Technology Institute of Machine and Materials Technology Gdańsk University of Technology Gdańsk Poland Electrode Design for Electrochemical Energy Systems University of Bayreuth Bayreuth Germany School of Chemistry University of New South Wales Sydney Australia School of Advanced Energy Sun Yat-Sen University Shenzhen China Sustainable Energy Center CSEM Neuchâtel2002 Switzerland Interdisciplinary Faculty of Science and Engineering Shimane University Matsue Japan Centre for Electronic and Optical Materials National Institute for Materials Science Tsukuba Japan Department of Civil and Industrial Engineering University of Pisa Pisa Italy Department of Materials Imperial College London Exhibition Road LondonSW7 2AZ United Kingdom Department of Chemical Engineering Massachusetts Institute of Technology Cambridge02139 United States Electrical Energy Systems University of Bayreuth Bayreuth Germany Ruder Boskovic Institute Zagreb Croatia Department of Materials and Minerals Resources Engineering National Taipei University Taipei Taiwan Finland Johann Radon Institute for Computational and Applied Mathematics Linz Austria School of Science and Engineering Nanjing Tech. University Nanjing China Karlsruhe Germany Department of Materials Science and Engineering University of Tennessee KnoxvilleTN37996 United States Physical Science Division Pacific Northwest National Laboratory RichlandWA99354 United States Systems Engineering for Electrical Energy Storage University of B

Electrochemical impedance spectroscopy (EIS) is widely used in electrochemistry, energy sciences, biology, and beyond. Analyzing EIS data is crucial, but it often poses challenges because of the numerous possible equivalent circuit models, the need for accurate analytical models, the difficulties of nonlinear regression, and the necessity of managing large datasets within a unified framework. To overcome these challenges, non-parametric models, such as the distribution of relaxation times (DRT, also known as the distribution function of relaxation times, DFRT), have emerged as promising tools for EIS analysis. For example, the DRT can be used to generate equivalent circuit models, initialize regression parameters, provide a time-domain representation of EIS spectra, and identify electrochemical processes. However, mastering the DRT method poses challenges as it requires mathematical and programming proficiency, which may extend beyond experimentalists’ usual expertise. Post-inversion analysis of DRT data can be difficult, especially in accurately identifying electrochemical processes, leading to results that may not always meet expectations. This article examines non-parametric EIS analysis methods, outlining their strengths and limitations from theoretical, computational, and end-user perspectives, and provides guidelines for their future development. Moreover, insights from survey data emphasize the need to develop a large impedance database, akin to an impedance genome. In turn, software development should target one-click, fully automated DRT analysis for multidimensional EIS spectra interpretation, software validation, and reliability. Particularly, creating a collaborative ecosystem hinged on free software could promote innovation and catalyze the adoption of the DRT method throughout all fields that use impedance data. © 2024 Elsevier Inc.

关键词： Electrochemical impedance spectroscopy

A Graph Neural Network Simulation of Dispersed Systems

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Hashemi, Aref Izadkhah, Aliakbar Department of Applied and Computational Mathematics and Statistics University of Notre Dame Notre DameIN United States Courant Institute New York University New YorkNY United States Department of Chemical Engineering Carnegie Mellon University PittsburghPA United States

We present a Graph Neural Network (GNN) that accurately simulates a multidisperse suspension of interacting spherical particles. Our machine learning framework is built upon the recent work of Sanchez-Gonzalez et al. ICML, PMLR, 119, 8459–8468 (2020) on graph network simulators, and efficiently learns the intricate dynamics of the interacting particles. Nodes and edges of the GNN correspond, respectively, to the particles with their individual properties/data (e.g., radius, position, velocity) and the pairwise interactions between the particles (e.g., electrostatics, hydrodynamics). A key contribution of our work is to account for the finite dimensions of the particles and their impact on the system dynamics. We test our GNN against a representative case study of a multidisperse mixture of two-dimensional spheres sedimenting under gravity in a liquid and interacting with each other by a Lennard-Jones potential. The present GNN framework offers a fast and accurate method for the theoretical study of complex physical systems such as field-induced behavior of colloidal suspensions and ionic liquids. Our implementation of the GNN is available on GitHub at ***/rfjd/GNS-DispersedSystems. Copyright © 2024, The Authors. All rights reserved.

关键词： Graph neural networks

Static Subspace Approximation for Random Phase Approximation Correlation Energies: Implementation and Performance

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Weinberg, Daniel Hull, Olivia A. Clary, Jacob M. Sundararaman, Ravishankar Vigil-Fowler, Derek Del Ben, Mauro Applied Mathematics & Computational Research Division Lawrence Berkeley National Laboratory BerkeleyCA United States Materials Chemical and Computational Science Directorate National Renewable Energy Laboratory GoldenCO United States Department of Materials Science and Engineering Rensselaer Polytechnic Institute TroyNY United States

Developing theoretical understanding of complex reactions and processes at interfaces requires using methods that go beyond semilocal density functional theory to accurately describe the interactions between solvent, reactants and substrates. Methods based on many-body perturbation theory, such as the random phase approximation (RPA), have previously been limited due to their computational complexity. However, this is now a surmountable barrier due to the advances in computational power available, in particular through modern GPU-based supercomputers. In this work, we describe the implementation of RPA calculations within BerkeleyGW and show its favorable computational performance on large complex systems relevant for catalysis and electrochemistry applications. Our implementation builds off of the static subspace approximation which, by employing a compressed representation of the frequency dependent polarizability, enables the evaluation of the RPA correlation energy with significant acceleration and systematically controllable accuracy. We find that the computational cost of calculating the RPA correlation energy scales only linearly with system size for systems containing up to 50 thousand bands, and is expected to scale quadratically thereafter. We also show excellent strong scaling results across several supercomputers, demonstrating the performance and portability of this implementation. Copyright © 2024, The Authors. All rights reserved.

关键词： Supercomputers