Particle Stochastic Dual Coordinate Ascent: Exponential convergent algorithm for mean field neural network optimization.
Kazusato OkoTaiji SuzukiAtsushi NitandaDenny WuPublished in: ICLR (2022)
Keyphrases
- coordinate ascent
- linear complexity
- optimization algorithm
- neural network
- global convergence
- k means
- image recovery
- optimization method
- cost function
- np hard
- particle swarm
- hybrid algorithm
- pattern recognition
- em algorithm
- global optimization
- objective function
- optimal solution
- higher dimensional
- global optimum
- distance transform
- back propagation
- particle swarm optimization
- optimization problems
- probabilistic model
- feature space