Non-asymptotic confidence bounds for stochastic approximation algorithms with constant step size (Q753365)
From MaRDI portal
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | Non-asymptotic confidence bounds for stochastic approximation algorithms with constant step size |
scientific article |
Statements
Non-asymptotic confidence bounds for stochastic approximation algorithms with constant step size (English)
0 references
1990
0 references
The multidimensional Robbins-Monro procedure with independent errors and constant step-size is considered. A comparison result for the stationary distribution of Markovian processes is derived which allows to bound the level of confidence regions. A stopping time which is based on the inner products of subsequent gradient estimates is introduced. It is shown that this stoppig time leads to confidence regions with a coverage probability which is uniform in the starting value.
0 references
non-asymptotic confidence bounds
0 references
multidimensional Robbins-Monro procedure
0 references
independent errors
0 references
constant step-size
0 references
comparison result
0 references
stationary distribution of Markovian processes
0 references
confidence regions
0 references
stopping time
0 references
inner products of subsequent gradient estimates
0 references
coverage probability
0 references
0 references
0 references