EUVIMEDEUVIMED entdecken
Crossref

A Recursively Recurrent Neural Network (R2N2) Architecture for Learning Iterative Algorithms

Danimir T. Doncevic, Alexander Mitsos, Yue Guo, Qianxiao Li, Felix Dietrich, Manuel Dahmen, Ioannis G. Kevrekidis

SIAM Journal on Scientific Computing · 30.04.2024 · 46 (2) · A719-A743

DOI: 10.1137/22m1535310 ↗

Kurz gefasst

Fachartikel in der Zeitschrift SIAM Journal on Scientific Computing, online erschienen am 06.03.2024. Verfasst von Danimir T. Doncevic und 6 weiteren Autor:innen. Die Arbeit wurde bisher 13-mal zitiert (Crossref) und nennt 92 Literaturangaben. Gefördert von: Air Force Office of Scientific Research, Deutsche Forschungsgemeinschaft, Helmholtz Association.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Metalearning of numerical algorithms for a given task consists of the data-driven identification and adaptation of an algorithmic structure and the associated hyperparameters. To limit the complexity of the metalearning problem, neural architectures with a certain inductive bias towards favorable algorithmic structures can, and should, be used. We generalize our previously introduced Runge–Kutta neural network to a recursively recurrent neural network superstructure for the design of customized iterative algorithms. In contrast to off-the-shelf deep learning approaches, it features a distinct division into modules for generation of information and for the subsequent assembly of this information towards a solution. Local information in the form of a subspace is generated by subordinate, inner, iterations of recurrent function evaluations starting at the current outer iterate. The update to the next outer iterate is computed as a linear combination of these evaluations, reducing the residual in this space, and constitutes the output of the network. We demonstrate that regular training of the weight parameters inside the proposed superstructure on input/output data of various computational problem classes yields iterations similar to Krylov solvers for linear equation systems, Newton–Krylov solvers for nonlinear equation systems, and Runge–Kutta integrators for ordinary differential equations. Due to its modularity, the superstructure can be readily extended with functionalities needed to represent more general classes of iterative algorithms traditionally based on Taylor series expansions.

Förderung

  • Air Force Office of Scientific Research
  • Deutsche Forschungsgemeinschaft
  • Helmholtz Association
  • National Research Foundation Singapore

Zitieren

APA

Doncevic, D. T., Mitsos, A., Guo, Y., Li, Q., Dietrich, F., Dahmen, M., & Kevrekidis, I. G. (2024). A Recursively Recurrent Neural Network (R2N2) Architecture for Learning Iterative Algorithms. SIAM Journal on Scientific Computing, 46(2), A719-A743. https://doi.org/10.1137/22m1535310

Vancouver

Doncevic DT, Mitsos A, Guo Y, Li Q, Dietrich F, Dahmen M, et al. A Recursively Recurrent Neural Network (R2N2) Architecture for Learning Iterative Algorithms. SIAM Journal on Scientific Computing. 2024;46(2):A719-A743. doi:10.1137/22m1535310

BibTeX
@article{Doncevic2024,
  author = {Doncevic, Danimir T. and Mitsos, Alexander and Guo, Yue and Li, Qianxiao and Dietrich, Felix and Dahmen, Manuel and Kevrekidis, Ioannis G.},
  title = {A Recursively Recurrent Neural Network (R2N2) Architecture for Learning Iterative Algorithms},
  journal = {SIAM Journal on Scientific Computing},
  year = {2024},
  volume = {46},
  number = {2},
  pages = {A719--A743},
  doi = {10.1137/22m1535310}
}
Kostenlos starten →