Multi-stage stochastic gradient method with momentum acceleration.
Zhijian LuoSiyu ChenYuntao QianYueen HouPublished in: Signal Process. (2021)
Keyphrases
- multistage
- gradient method
- stochastic programming
- stochastic optimization
- convergence rate
- learning rate
- assembly systems
- production system
- step size
- single stage
- optimization methods
- negative matrix factorization
- lot sizing
- dynamic programming
- natural gradient learning
- topic models
- optimal policy
- supply chain
- cost function
- lot streaming
- multiscale