Catalogue Search | MBRL
Are you sure you want to remove the book from the shelf?
{{itemTitle}}
115,644
result(s) for
"Computational mathematics"
Sort by:
Error analysis for physics-informed neural networks (PINNs) approximating Kolmogorov PDEs
Physics-informed neural networks approximate solutions of PDEs by minimizing pointwise residuals. We derive rigorous bounds on the error, incurred by PINNs in approximating the solutions of a large class of linear parabolic PDEs, namely Kolmogorov equations that include the heat equation and Black-Scholes equation of option pricing, as examples. We construct neural networks, whose PINN residual (generalization error) can be made as small as desired. We also prove that the total L2-error can be bounded by the generalization error, which in turn is bounded in terms of the training error, provided that a sufficient number of randomly chosen training (collocation) points is used. Moreover, we prove that the size of the PINNs and the number of training samples only grow polynomially with the underlying dimension, enabling PINNs to overcome the curse of dimensionality in this context. These results enable us to provide a comprehensive error analysis for PINNs in approximating Kolmogorov PDEs.
Journal Article
Stability and error estimates of the SAV Fourier-spectral method for the phase field crystal equation
We consider fully discrete schemes based on the scalar auxiliary variable (SAV) approach and stabilized SAV approach in time and the Fourier-spectral method in space for the phase field crystal (PFC) equation. Unconditionally, energy stability is established for both first- and second-order fully discrete schemes. In addition to the stability, we also provide a rigorous error estimate which shows that our second-order in time with Fourier-spectral method in space converges with order O(Δt2 + N−m), where Δt, N, and m are time step size, number of Fourier modes in each direction, and regularity index in space, respectively. We also present numerical experiments to verify our theoretical results and demonstrate the robustness and accuracy of the schemes.
Journal Article
Wonders beyond numbers : a brief history of all things mathematical
Running in something approaching chronological order, this book shows that every breakthrough in math represents a single step forward, resting on the work of others, and brings to life the importance of numbers, shapes, and patterns in the world around us.
Book
Neural network closures for nonlinear model order reduction
Many reduced-order models are neither robust with respect to parameter changes nor cost-effective enough for handling the nonlinear dependence of complex dynamical systems. In this study, we put forth a robust machine learning framework for projection-based reduced-order modeling of such nonlinear and nonstationary systems. As a demonstration, we focus on a nonlinear advection-diffusion system given by the viscous Burgers equation, which is a prototypical setting of more realistic fluid dynamics applications due to its quadratic nonlinearity. In our proposed methodology the effects of truncated modes are modeled using a single layer feed-forward neural network architecture. The neural network architecture is trained by utilizing both the Bayesian regularization and extreme learning machine approaches, where the latter one is found to be more computationally efficient. A significant emphasis is laid on the selection of basis functions through the use of both Fourier bases and proper orthogonal decomposition. It is shown that the proposed model yields significant improvements in accuracy over the standard Galerkin projection methodology with a negligibly small computational overhead and provide reliable predictions with respect to parameter changes.
Journal Article
A reduced order variational multiscale approach for turbulent flows
The purpose of this work is to present different reduced order model strategies starting from full order simulations stabilized using a residual-based variational multiscale (VMS) approach. The focus is on flows with moderately high Reynolds numbers. The reduced order models (ROMs) presented in this manuscript are based on a POD-Galerkin approach. Two different reduced order models are presented, which differ on the stabilization used during the Galerkin projection. In the first case, the VMS stabilization method is used at both the full order and the reduced order levels. In the second case, the VMS stabilization is used only at the full order level, while the projection of the standard Navier-Stokes equations is performed instead at the reduced order level. The former method is denoted as consistent ROM, while the latter is named non-consistent ROM, in order to underline the different choices made at the two levels. Particular attention is also devoted to the role of inf-sup stabilization by means of supremizers in ROMs based on a VMS formulation. Finally, the developed methods are tested on a numerical benchmark.
Journal Article
Mathematical modeling for intelligent systems : theory, methods, and simulation
\"Mathematical Modeling for Intelligent Systems: Theory, Methods and Simulation aims to provide a reference for the applications of mathematical modeling using intelligent techniques in various unique industry problems in the era of Industry 4.0. Providing a thorough introduction to the field of soft computing techniques, the book covers every major technique in artificial intelligence in a clear and practical style. It also highlights current research and applications, addresses issues encountered in the development of applied systems, and describes a wide range of intelligent systems techniques, including neural networks, fuzzy logic, evolutionary strategy, and genetic algorithms. The book demonstrates concepts through simulation examples and practical experimental results. The book offers a well-balanced mathematical analysis of modelling physical systems. Summarizes basic principles in differential geometry and convex analysis as needed. The book covers a wide range of industrial and social applications, and bridges the gap between core theory and costly experiments through simulations and modelling. The focus of the book is manifold ranging from stability of fluid flows, nano fluids, drug delivery, and security of image data to Pandemic modeling etc. The book is primarily aimed at advanced undergraduates and postgraduate students studying computer science, mathematics and statistics. Researchers and professionals will also find this book useful\"-- Provided by publisher.
BOOK
Randomized algorithms for the approximations of Tucker and the tensor train decompositions
Randomized algorithms provide a powerful tool for scientific computing. Compared with standard deterministic algorithms, randomized algorithms are often faster and robust. The main purpose of this paper is to design adaptive randomized algorithms for computing the approximate tensor decompositions. We give an adaptive randomized algorithm for the computation of a low multilinear rank approximation of the tensors with unknown multilinear rank and analyze its probabilistic error bound under certain assumptions. Finally, we design an adaptive randomized algorithm for computing the tensor train approximations of the tensors. Based on the bounds about the singular values of sub-Gaussian matrices with independent columns or independent rows, we analyze these randomized algorithms. We illustrate our adaptive randomized algorithms via several numerical examples.
Journal Article