Two Recurrent Neural Networks With Reduced Model Complexity for Constrained l(1)-Norm Optimization

Youshen Xia, Jun Wang, Zhenyu Lu, Liqing Huang

Research output: Contribution to journalArticlepeer-review

Abstract

Because of the robustness and sparsity performance of least absolute deviation (LAD or l₁) optimization, developing effective solution methods becomes an important topic. Recurrent neural networks (RNNs) are reported to be capable of effectively solving constrained l₁-norm optimization problems, but their convergence speed is limited. To accelerate the convergence, this article introduces two RNNs, in form of continuous- and discrete-time systems, for solving l₁-norm optimization problems with linear equality and inequality constraints. The RNNs are theoretically proven to be globally convergent to optimal solutions without any condition. With reduced model complexity, the two RNNs can significantly expedite constrained l₁-norm optimization. Numerical simulation results show that the two RNNs spend much less computational time than related RNNs and numerical optimization algorithms for linearly constrained l₁-norm optimization.
Original languageEnglish
JournalIEEE Transactions on Neural Networks and Learning Systems
DOIs
Publication statusPublished - 2022
Externally publishedYes

Keywords

  • Approximation algorithms
  • Complexity theory
  • Computational modeling
  • Convergence
  • Fast computation
  • linearly constrained l₁-norm optimization
  • Mathematical models
  • model complexity
  • Optimization
  • recurrent neural network (RNN).
  • Recurrent neural networks

Fingerprint

Dive into the research topics of 'Two Recurrent Neural Networks With Reduced Model Complexity for Constrained l(1)-Norm Optimization'. Together they form a unique fingerprint.

Cite this