Global ETD Search

Search theses and dissertations gathered from participating repositories worldwide. Every result links back to the library that holds it. No account is needed.

Results

Showing 1 to 5 of 5 for “"Schwarz Methods"”.

  1. Hessian Matrix-Free Lagrange-Newton-Krylov-Schur-Schwarz Methods for Elliptic Inverse Problems

    … in measurements of the data. Regularization methods are employed to partially alleviate this problem. The PDE-constrained optimization is formulated as the minimization of a Lagrangian functional, formed from the regularized cost function and the discretized PDE, with respect to the …

    odu Repository record for Hessian Matrix-Free Lagrange-Newton-Krylov-Schur-Schwarz Methods for Elliptic Inverse Problems (opens in a new tab)

  2. Domain decomposition methods for time-harmonic elastic waves

    … difficult to solve both by direct and iterative methods. Domain decomposition methods are hybrid methods in the sense that they use an iterative coupling of smaller problems that are solved by direct methods,and rely on the splitting the global problem into local problems on smaller subdomains. …

    strathclyde Repository record for Domain decomposition methods for time-harmonic elastic waves (opens in a new tab)

  3. From efficient high-order methods to scientific machine learning.

    … some limitations of the current Finite Element Methods (FEM) packages. We focus on the efficient implementation of implicit time discretization schemes, achieving computational savings by exploiting the structure of serendipity elements and developing a wrapper that integrates Machine Learning …

    tdl Repository record for From efficient high-order methods to scientific machine learning. (opens in a new tab)

  4. From efficient high-order methods to scientific machine learning.

    … some limitations of the current Finite Element Methods (FEM) packages. We focus on the efficient implementation of implicit time discretization schemes, achieving computational savings by exploiting the structure of serendipity elements and developing a wrapper that integrates Machine Learning …

    baylor Repository record for From efficient high-order methods to scientific machine learning. (opens in a new tab)

  5. Exascale finite/spectral element simulations

    Submission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2026-08-01

    uiuc Repository record for Exascale finite/spectral element simulations (opens in a new tab)