Recurrent Neural Network for Convex Optimization Problems with Application in Robotics

来源 :第八届工业与应用数学国际大会 | 被引量 : 0次 | 上传用户:ZHANGXIANYU0000
下载到本地 , 更方便阅读
声明 : 本文档内容版权归属内容提供方 , 如果您对本文有版权争议 , 可与客服联系进行内容授权或下架
论文部分内容阅读
  In this talk,we introduce a recurrent neural network model which is capable of solving the convex optimization problems.Both the objective function and inequality constraints can be non-smooth.
其他文献
  Conic integral geometry provides deep insights as well as-exactformulas for the probabilities of important events such as-success of the compressed sensing
会议
  This min-symposium considers a variety of ill-posed inverse problems associated with information theory,signal processing,and image processing.
会议
  We present a parallel subspace decomposition framework for nonlinear optimization,which can be regarded as an extension of the domain decomposition method f
会议
  Direct-search methods are a class of derivative-free algorithms based on evaluating the objective function along directions in positive spanning sets.
会议
  We present software that uses concurrent evaluations of the objective function to identify all of its local minima.The method is based on a multistart algor
会议
  Black-box optimization problems arise frequently in real-world applications.In these problems,the function values are provided by blackboxes(typically simul
会议
  We study the geometry and duality of multiple interacting heterogeneous multi-agent systems where the agent dynamics may not be the same.
会议
  This paper study stochastic control for systems with transmission delay.A sufficient and necessary stabilizing condition is presented for the stochastic sys
会议
  In this report,the problem of distributed output tracking for highorder uncertain nonlinear multi-agent systems is presented.For the case where the graph to
会议
  Measurement and communication uncertainties are important factors for the stability and performances of cooperative dynamic networks.
会议