Geometrical convergence rate for distributed optimization with time-varying directed graphs and uncoordinated step-sizes.
Qingguo LüHuaqing LiDawen XiaPublished in: Inf. Sci. (2018)
Keyphrases
- step size
- convergence rate
- directed graph
- global convergence
- faster convergence
- convergence speed
- line search
- faster convergence rate
- learning rate
- random walk
- evolutionary programming
- gradient method
- primal dual
- variable step size
- undirected graph
- conjugate gradient
- global optimum
- global search
- mutation operator
- numerical stability
- cost function
- directed acyclic graph
- particle swarm optimization algorithm
- optimization method
- global optimization
- evolutionary algorithm
- genetic programming
- objective function
- reinforcement learning
- number of iterations required
- special case
- optimization problems
- optimization algorithm
- gray scale
- particle swarm