The Experts below are selected from a list of 8910 Experts worldwide ranked by ideXlab platform
Uday V Shanbhag - One of the best experts on this subject based on the ideXlab platform.
-
regularized iterative stochastic approximation methods for stochastic variational inequality problems
IEEE Transactions on Automatic Control, 2013Co-Authors: Jayash Koshal, A Nedic, Uday V ShanbhagAbstract:We consider a Cartesian stochastic variational inequality problem with a monotone map. Monotone stochastic variational inequalities arise naturally, for instance, as the equilibrium conditions of monotone stochastic Nash games over continuous strategy sets or multiuser stochastic optimization problems. We introduce two classes of stochastic approximation methods, each of which requires exactly one projection step at every iteration, and provide convergence analysis for each of them. Of these, the first is a stochastic iterative Tikhonov regularization method which necessitates the update of the regularization parameter after every iteration. The second method is a stochastic iterative proximal-point method, where the centering term is updated after every iteration. The Cartesian structure lends itself to constructing distributed multi-agent extensions and conditions are provided for recovering global convergence in limited Coordination variants where agents are allowed to choose their steplength sequences, regularization and centering parameters independently, while meeting a suitable Coordination Requirement. We apply the proposed class of techniques and their limited Coordination versions to a stochastic networked rate allocation problem.
Jayash Koshal - One of the best experts on this subject based on the ideXlab platform.
-
regularized iterative stochastic approximation methods for stochastic variational inequality problems
IEEE Transactions on Automatic Control, 2013Co-Authors: Jayash Koshal, A Nedic, Uday V ShanbhagAbstract:We consider a Cartesian stochastic variational inequality problem with a monotone map. Monotone stochastic variational inequalities arise naturally, for instance, as the equilibrium conditions of monotone stochastic Nash games over continuous strategy sets or multiuser stochastic optimization problems. We introduce two classes of stochastic approximation methods, each of which requires exactly one projection step at every iteration, and provide convergence analysis for each of them. Of these, the first is a stochastic iterative Tikhonov regularization method which necessitates the update of the regularization parameter after every iteration. The second method is a stochastic iterative proximal-point method, where the centering term is updated after every iteration. The Cartesian structure lends itself to constructing distributed multi-agent extensions and conditions are provided for recovering global convergence in limited Coordination variants where agents are allowed to choose their steplength sequences, regularization and centering parameters independently, while meeting a suitable Coordination Requirement. We apply the proposed class of techniques and their limited Coordination versions to a stochastic networked rate allocation problem.
Junliang Chen - One of the best experts on this subject based on the ideXlab platform.
-
Declarative Construction of Distributed Event-driven IoT Services Based on IoT Resource Models
IEEE Transactions on Services Computing, 2017Co-Authors: Yang Zhang, Junliang ChenAbstract:In IoT (Internet of Things) scenarios, the Coordination of physical systems is often complex and rigid. Work to date has not comprehensively explored how to flexibly construct distributed IoT services to satisfy the Coordination Requirement. In our work, we propose a declarative approach to construct an event-driven IoT service system, where physical devices and systems are explicitly modeled as a service architecture foundation, the service behavior is flexibly declared based on distributed events, and it is step-by-step refined to have rigid service properties while allowing runtime adaptation. As an example, the controllability property of physical systems is discussed. Finally, we establish a distributed event-driven IoT service platform to test our method. Some experiments are made to concept-prove our work.
Dongjun Lee - One of the best experts on this subject based on the ideXlab platform.
-
passive decomposition of mechanical systems with Coordination Requirement
IEEE Transactions on Automatic Control, 2013Co-Authors: Dongjun LeeAbstract:We show the fundamental passive decomposition property of general mechanical systems on a n -dim. configuration manifold M, i.e., when endowed with a submersion h:M→N , where N is a m -dim. manifold (m ≤ n), their Lagrangian dynamics with the kinetic energy as the Lagrangian can always be decomposed into: 1) shape system, describing the m-dim. dynamics of h(q) on N ; 2) locked system, representing the (n-m)-dim. dynamics along the level set of h; and 3) energetically-conservative coupling between them. The locked and shape systems also individually inherit the Lagrangian structure and passivity of the original dynamics. We exhibit and analyze geometric and energetic properties of the passive decomposition in a coordinate-free manner. An illustrative example on SO(3) is also provided.
A Nedic - One of the best experts on this subject based on the ideXlab platform.
-
regularized iterative stochastic approximation methods for stochastic variational inequality problems
IEEE Transactions on Automatic Control, 2013Co-Authors: Jayash Koshal, A Nedic, Uday V ShanbhagAbstract:We consider a Cartesian stochastic variational inequality problem with a monotone map. Monotone stochastic variational inequalities arise naturally, for instance, as the equilibrium conditions of monotone stochastic Nash games over continuous strategy sets or multiuser stochastic optimization problems. We introduce two classes of stochastic approximation methods, each of which requires exactly one projection step at every iteration, and provide convergence analysis for each of them. Of these, the first is a stochastic iterative Tikhonov regularization method which necessitates the update of the regularization parameter after every iteration. The second method is a stochastic iterative proximal-point method, where the centering term is updated after every iteration. The Cartesian structure lends itself to constructing distributed multi-agent extensions and conditions are provided for recovering global convergence in limited Coordination variants where agents are allowed to choose their steplength sequences, regularization and centering parameters independently, while meeting a suitable Coordination Requirement. We apply the proposed class of techniques and their limited Coordination versions to a stochastic networked rate allocation problem.