A NOISE-TOLERANT AND FIXED-TIME ZEROING NEURAL NETWORK MODEL FOR SOLVING MULTI-LINEAR $ \mathcal{M} $-TENSOR EQUATIONS
Year: 2026
Type: article
Abstract: Zeroing neural network (ZNN) models are effective methods for solving multi-linear systems. To improve the performance of ZNN models, a new ZNN model is proposed for solving multi-linear systems with $ \mathcal{M} $-tensor in this paper. The proposed ZNN model incorporates with a nonlinear activation function. Convergence and robustness of the proposed ZNN model are analyzed. Theoretical results reveal that the proposed ZNN model converges within a fixed time and tolerates different types of noise. Numerical experiments illustrate that the proposed ZNN model is more powerful than some existing ZNN models for solving multi-linear systems with $ \mathcal{M} $-tensor.
DOI: 10.11948/20250257
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W7203840369