On the convergence properties of a K-step averaging stochastic gradient descent algorithm for nonconvex optimization.
Fan ZhouGuojing CongPublished in: CoRR (2017)
Keyphrases
- stochastic gradient descent
- stochastic gradient
- optimization algorithm
- learning algorithm
- optimal solution
- convergence rate
- kalman filter
- iterative algorithms
- worst case
- np hard
- loss function
- global optimization
- probabilistic model
- step size
- online algorithms
- cost function
- objective function
- online learning
- convex hull
- prior information