Because of the robustness and sparsity performance of least absolute deviation (LAD or l ) optimization, developing effective solution methods becomes an important topic. Recurrent neural networks (RNNs) are reported to be capable of effectively solving constrained l -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 l -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 l -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 l -norm optimization.
Download full-text PDF |
Source |
|---|---|
| http://dx.doi.org/10.1109/TNNLS.2021.3133836 | DOI Listing |
Publication Analysis
Top Keywords
Similar Publications
The Auditory Environment at Early Intervention Groups for Young Children With Hearing Loss: Signal to Noise Ratio, Background Noise, and Reverberation.
Ear Hear
January 2025
Dutch Foundation of the Deaf and Hard of Hearing Child (NSDSK), Amsterdam, The Netherlands.
Objectives: One important aspect in facilitating language access for children with hearing loss (HL) is the auditory environment. An optimal auditory environment is characterized by high signal to noise ratios (SNRs), low background noise levels, and low reverberation times. In this study, the authors describe the auditory environment of early intervention groups specifically equipped for young children with HL.
View Article and Find Full Text PDFAccurate image reconstruction within and beyond the field-of-view of CT system from data with truncation.
Phys Med Biol
January 2025
Department of Radiology, University of Chicago, 5841 South Maryland Avenue, Chicago, IL 60637, USA, Chicago, 60637, UNITED STATES.
Objective: Accurate image reconstruction from data with truncation in X-ray computed tomography (CT) remains a topic of research interest; and the works reported previously in the literature focus largely on reconstructing an image only within the scanning field-of-view (FOV). We develop algorithms to invert the data model with truncation for accurate image reconstruction within the entire subject support or a region slightly smaller than the subject support.
Methods: We formulate image reconstruction from data with truncation as an optimization program, which includes hybrid constraints on image total variation (TV) and image L1-norm for effectively suppressing truncation artifacts.
Multi-view clustering based on feature selection and semi-non-negative anchor graph factorization.
Neural Netw
January 2025
School of Mathematical Sciences, Harbin Engineering University, Harbin 150001, China.
Multi-view clustering has garnered significant attention due to its capacity to utilize information from multiple perspectives. The concept of anchor graph-based techniques was introduced to manage large-scale data better. However, current methods rely on K-means or uniform sampling to select anchors in the original space.
View Article and Find Full Text PDFC-parameter version of robust bounded one-class support vector classification.
Sci Rep
January 2025
College of Mathematics and Systems Science, Xinjiang University, Urumqi , 830046, China.
ν-one-class support vector classification (ν-OCSVC) has garnered significant attention for its remarkable performance in handling single-class classification and anomaly detection. Nonetheless, the model does not yield a unique decision boundary, and potentially compromises learning performance when the training data is contaminated by some outliers or mislabeled observations. This paper presents a novel C-parameter version of bounded one-class support vector classification (C-BOCSVC) to determine a unique decision boundary.
View Article and Find Full Text PDFDistributed aggregative optimization with affine coupling constraints.
Neural Netw
December 2024
Shanghai Research Institute for Intelligent Autonomous Systems, Tongji University, Shanghai 201210, PR China; Department of Control Science and Engineering, College of Electronics and Information Engineering, Tongji University, Shanghai 201804, PR China; National Key Laboratory of Autonomous Intelligent Unmanned Systems, Shanghai, PR China; Frontiers Science Center for Intelligent Autonomous Systems, Ministry of Education, Shanghai, PR China. Electronic address:
This paper investigates a distributed aggregative optimization problem subject to coupling affine inequality constraints, in which local objective functions depend not only on their own decision variables but also on an aggregation of all the agents' variables. The formulated problem encompasses numerous practical applications, such as commodity distribution, electric vehicle charging, and energy consumption control in power grids. Hence, there is a compelling need to explore a new neurodynamic approach to address this.
View Article and Find Full Text PDFWant AI Summaries of new PubMed Abstracts delivered to your In-box?
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!