Distributed and Stochastic Optimization Methods with Gradient Compression and Local Steps.
Eduard GorbunovPublished in: CoRR (2021)
Keyphrases
- optimization methods
- stochastic methods
- gradient method
- stage stochastic programs
- optimization method
- optimization problems
- simulated annealing
- unconstrained optimization
- optimization approaches
- distributed environment
- distributed systems
- global convergence
- continuous optimization
- quasi newton
- compression algorithm
- compression ratio
- particle swarm
- data compression
- convex relaxation
- bayesian network models
- direct optimization
- image compression
- compression scheme
- evolution strategy
- multi agent
- peer to peer
- trust region
- objective function
- edge detection
- efficient optimization
- search algorithm
- global optimum
- evolutionary algorithm
- monte carlo
- mobile agents