60,865 Hits in 3.6 sec

A Novel Recurrent Neural Network for Solving Nonlinear Optimization Problems With Inequality Constraints

Youshen Xia, Gang Feng, Jun Wang
2008 IEEE Transactions on Neural Networks  
Compared with variety of the existing projection neural networks, including their extensions and modification, for solving such nonlinearly constrained optimization problems, it is shown that the proposed  ...  This paper presents a novel recurrent neural network for solving nonlinear optimization problems with inequality constraints.  ...  The extended projection neural network models for solving (1) can be described by (10) It is easy to see that the extended projection neural network model (10) has the same network complexity with the  ... 
doi:10.1109/tnn.2008.2000273 pmid:18701366 fatcat:njhz6u63mrcstbbplkjr2yuiby

A projection neural network and its application to constrained optimization problems

Youshen Xia, H. Leung, Jun Wang
2002 IEEE Transactions on Circuits and Systems I Fundamental Theory and Applications  
In this paper, we present a recurrent neural network for solving the nonlinear projection formulation.  ...  Index Terms-Constrained optimization problems, global stability, recurrent neural network.  ...  ACKNOWLEDGMENT The authors would like thank the reviewers for offering valuable suggestions and comments leading to an improvement in this paper.  ... 
doi:10.1109/81.995659 fatcat:cwg5rpe6frhf5o754rqqhoz4qm

Robust Linear Neural Network for Constrained Quadratic Optimization

Zixin Liu, Yuanan Liu, Lianglin Xiong
2017 Discrete Dynamics in Nature and Society  
Based on the feature of projection operator under box constraint, by using convex analysis method, this paper proposed three robust linear systems to solve a class of quadratic optimization problems.  ...  Finally, a numerical simulation example and an application example in compressed sensing problem are also given to illustrate the validity of the criteria established in this paper.  ...  These new applications in robotics extend new application fields for projection neural network.  ... 
doi:10.1155/2017/5073640 fatcat:hfsa7l67cbhr3cvifg2akavmjy

A dual neural network for convex quadratic programming subject to linear equality and inequality constraints

Yunong Zhang, Jun Wang
2002 Physics Letters A  
A recurrent neural network called the dual neural network is proposed in this Letter for solving the strictly convex quadratic programming problems.  ...  The global convergence behavior of the dual neural network is demonstrated by an illustrative numerical example.   ...  One approach commonly used in developing an optimization neural network is to first convert the constrained optimization problem into an associated unconstrained optimization problem, and then design a  ... 
doi:10.1016/s0375-9601(02)00424-3 fatcat:g6drzblej5gmnltuyfabap4nnm

FISAR: Forward Invariant Safe Reinforcement Learning with a Deep Neural Network-Based Optimize [article]

Chuangchuang Sun, Dong-Ki Kim, Jonathan P. How
2021 arXiv   pre-print
To address this, we propose to learn a generic deep neural network (DNN)-based optimizer to optimize the objective while satisfying the linear constraints.  ...  To the best of our knowledge, this is the first DNN-based optimizer for constrained optimization with the forward invariance guarantee.  ...  APPENDIX The hyperparameters for our method and the baselines can be found in the following table, where "NN" and "lr" stand for "neural network" and "learning rate", respectively.  ... 
arXiv:2006.11419v4 fatcat:x3qzrb2wtve6hoplr76c6uhzii

A Recurrent Neural Network for Solving Nonlinear Convex Programs Subject to Linear Constraints

Y. Xia, J. Wang
2005 IEEE Transactions on Neural Networks  
It is shown here that the proposed neural network is stable in the sense of Lyapunov and globally convergent to an optimal solution within a finite time under the condition that the objective function  ...  The proposed neural network has a simpler structure and a lower complexity for implementation than the existing neural networks for solving such problems.  ...  In addition, several projection neural networks for constrained optimization and related problems were developed [11] , [19] .  ... 
doi:10.1109/tnn.2004.841779 pmid:17385632 fatcat:dcvgi32m7bcz3ows237zpdgoz4

A Recurrent Neural Network for Nonlinear Convex Optimization Subject to Nonlinear Inequality Constraints

Y. Xia, J. Wang
2004 IEEE Transactions on Circuits and Systems I Fundamental Theory and Applications  
Simulation results are given to illustrate further the global convergence and performance of the proposed neural network for constrained nonlinear optimization.  ...  Compared with the existing neural networks for solving such nonlinear optimization problems, the proposed neural network has two major advantages.  ...  By using the projection technique for constrained optimization, an equivalent relationship between a NCP and a set of projection formulations is established.  ... 
doi:10.1109/tcsi.2004.830694 fatcat:ljzby24izrgrpkpxzmhqqfj3bu

Second-Order Networks in PyTorch [chapter]

Daniel Brooks, Olivier Schwander, Frédéric Barbaresco, Jean-Yves Schneider, Matthieu Cord
2019 Lecture Notes in Computer Science  
In this work we propose a Python library which implements neural networks on SPD matrices, based on the popular deep learning framework Pytorch.  ...  For an image recognition task, these features may come from a pre-trained deep network but nothing keeps from training the whole network in an end-to-end fashion or to fine-tune the parameters.  ...  In the following section we describe the core components of a SPD neural network, which we may call SPDNet. The third section deals with the optimization of a manifold-valued network.  ... 
doi:10.1007/978-3-030-26980-7_78 fatcat:idkushgvqrehnjurqw2dc6hubq

Solving convex optimization problems using recurrent neural networks in finite time

Long Cheng, Zeng-Guang Hou, Noriyasu Homma, Min Tan, Madam M. Gupta
2009 2009 International Joint Conference on Neural Networks  
Moreover, by using the penalty function method, the proposed neural network can be extended straightforwardly to solving the constrained optimization problem.  ...  A recurrent neural network is proposed to deal with the convex optimization problem.  ...  Furthermore, by using the penalty function method, the proposed neural network can be extended straightforward ly for the constrained optimization case.  ... 
doi:10.1109/ijcnn.2009.5178723 dblp:conf/ijcnn/ChengHHTG09 fatcat:pxj3bkd4f5h3vgrg36224nr4hm

Learning Constrained Distributions of Robot Configurations with Generative Adversarial Network [article]

Teguh Santoso Lembono, Emmanuel Pignat, Julius Jankowski, Sylvain Calinon
2021 arXiv   pre-print
Then, we use it to generate samples in sampling-based constrained motion planning algorithms to reduce the necessary projection steps, speeding up the computation.  ...  We propose a generative adversarial network approach to learn the distribution of valid robot configurations under such constraints.  ...  The generator consists of an ensemble of Nnet neural networks, while the discriminator consists of a single neural network.  ... 
arXiv:2011.05717v2 fatcat:c5bz35z2qvc5zhbhe6yo5ofhfy

Optimal information loading into working memory in prefrontal cortex [article]

Jake P Stroud, Kei Watanabe, Takafumi Suzuki, Mark G Stokes, Máté Lengyel
2021 bioRxiv   pre-print
However, how information is loaded into working memory for subsequent maintenance remains poorly understood.  ...  The neural circuit mechanisms underlying this information maintenance are thought to rely on persistent activities resulting from attractor dynamics.  ...  We thank Flavia Mancini, John Duncan, Guillaume Hennequin, and Yashar Ahmadian for useful feedback and detailed comments on the manuscript.  ... 
doi:10.1101/2021.11.16.468360 fatcat:jpixppojk5ab5hz3pmhb4xwygm

Solving Pseudomonotone Variational Inequalities and Pseudoconvex Optimization Problems Using the Projection Neural Network

Xiaolin Hu, Jun Wang
2006 IEEE Transactions on Neural Networks  
In recent years, a recurrent neural network called projection neural network was proposed for solving monotone variational inequalities and related convex optimization problems.  ...  Since monotonicity is a special case of pseudomononicity, the projection neural network can be applied to solve a broader class of constrained optimization problems related to variational inequalities.  ...  CONCLUDING REMARKS In this paper, we extended the scope of an existing projection neural network, which was originally proposed for solving monotone variational inequalities, to pseudomonotone variational  ... 
doi:10.1109/tnn.2006.879774 pmid:17131663 fatcat:acgv5m7rkbfrljdvxa6carmwsu

Constrained Motion Planning Networks X [article]

Ahmed H. Qureshi, Jiangeng Dong, Asfiya Baig, Michael C. Yip
2021 arXiv   pre-print
To address this challenge, we present Constrained Motion Planning Networks X (CoMPNetX).  ...  It is a neural planning approach, comprising a conditional deep neural generator and discriminator with neural gradients-based fast projection operator.  ...  ACKNOWLEDGMENTS We thank Dmitry Berenson and Frank Park for their insightful discussions and sharing their algorithms' implementations.  ... 
arXiv:2010.08707v2 fatcat:rd2drgc5jvgyflwuf24mazzafy

A Novel Recurrent Adaptive Backstepping Optimal Control Strategy for a Single Inverted Pendulum System [article]

Mohammad Sarbaz
2021 arXiv   pre-print
By this method, an inverted pendulum is stabilized using projection recurrent neural network-based adaptive backstepping control (PRNN-ABC).  ...  To study the recurrent neural network (RNN) according to the Karush- Kuhn-Tucker (KKT) optimization conditions and the variational inequality, the dynamic model of the RNN will be derived.  ...  Therefore, to solve this problem, an optimal backstepping control approach based on projection recurrent neural network is studied in this work for inverted pendulum.  ... 
arXiv:2110.09846v1 fatcat:lvulahp355cx7a23gjimodepey

Table of contents

2018 IEEE Transactions on Neural Networks and Learning Systems  
Zhang Runtime Programmable and Memory Bandwidth Optimized FPGA-Based Coprocessor for Deep Convolutional Neural Network ............................................................................ N.  ...  Chen 6113 Kurtosis-Based CRTRL Algorithms for Fully Connected Recurrent Neural Networks .... E. C. Mengüç and N.  ... 
doi:10.1109/tnnls.2018.2880596 fatcat:bb7s44kagzcr7ecchczfxzznhi
« Previous Showing results 1 — 15 out of 60,865 results