On extensions of Polyak's averaging approach to stochastic approximation

@article{Yin1991OnEO,
  title={On extensions of Polyak's averaging approach to stochastic approximation},
  author={George Yin},
  journal={Stochastics and Stochastics Reports},
  year={1991},
  volume={36},
  pages={245-264}
}
  • G. Yin
  • Published 1 September 1991
  • Mathematics, Computer Science
  • Stochastics and Stochastics Reports
Stochastic approximation algorithms are considered and Polyak's averaging approach (cf. [1]) is revisited. Under much weaker conditions, convergence and rate of convergence results are developed. In lieu of uncorrelated noise, φ-mixing type of random disturbances are treated. By means of weak convergence methods, it is shown that the multistage algorithms via averaging have asymptotically optimal convergence speed and are efficient procedures 
Asymptotically optimal rate of convergence of smoothed stochastic recursive algorithms
TLDR
Under correlated noise processes, it is shown that a scaled sequence of the iterates converges weakly to a Browman motion, which means the algorithm is asymptotically optimal.
Asymptotically efficient stochastic approximation
Asymptotic efficiency of a stochastic approximation algorithm is proved under weak conditions: no growth rate restriction is imposed on the regression function; wide class of observation noises is
On Stochastic Approximation Procedures with Averaging
Linear multidimensional stochastic approximation procedures in continuous time with martingale errors are considered. An asymptotic behavior of the estimator obtained by trajectory averaging is
Weighted averaging and stochastic approximation
TLDR
It is proved that the convergence of a stochastic approximation algorithm is equivalent to the convergence to the weighted average of the associated noise sequence, and it is shown that the averaged stochastics approximation algorithms can tolerate a larger class of noise sequences than the stand-alone stochastically approximation algorithms.
Weighted averaging and stochastic approximation
TLDR
It is proved that the convergence of a stochastic approximation algorithm is equivalent to the convergence to the weighted average of the associated noise sequence, and it is shown that the averaged stochastics approximation algorithms can tolerate a larger class of noise sequences than the stand-alone stochastically approximation algorithms.
About Gaussian schemes in stochastic approximation
Continuous-time stochastic approximation: Convergence and asymptotic efficiency
A continuous-time stochastic approximation algorithm is proposed. It is shown that the estimate x y is strongly consistent and the averaged estimate is asymptotically efficient. The characteristics
The Robbins-Monro type stochastic differential equations. III. Polyak's averaging
General results concerning the asymptotic behaviour of the Polyak averaging of the solution of the Robbins-Monro type stochastic differential equation are presented. It is shown that the suitable
On a continuous time stochastic approximation problem
TLDR
The main effort is to derive the asymptotic properties of the algorithm, and it is shown that ast → ∞, a suitably normalized sequence of the estimation error,Τ√t(¯xtr−θ) is equivalent to a scaled sequences of the random noise process, namely, (1/ ∼t)∫0tr ξsds.
...
...

References

SHOWING 1-10 OF 10 REFERENCES
Asymptotic properties of distributed and communication stochastic approximation algorithms
The asymptotic properties of extensions of the type of distributed or decentralized stochastic approximation proposed in [1] are developed. Such algorithms have numerous potential applications in
Stochastic approximation algorithms for parallel and distributed processing
TLDR
An interesting class of "distributed" recursive stochastic algorithms that arises when parallel processing methods are used for the Monte Carlo optimization of systems, as well as in applications such as decentralized and asynchronous on-line optimization of the flows in communication networks are treated.
Approximation and Weak Convergence Methods for Random Processes
Control and communications engineers, physicists, and probability theorists, among others, will find this book unique. It contains a detailed development of approximation and limit theorems and
Consistency and asymptotic efficiency of slope estimates in stochastic approximation schemes
Adaptive stochastic approximation schemes for choosing the levels of x at which y is to be observed are useful in applications of the following nature. Suppose that in (1.1) x~ is the dosage level of
Adaptive Algorithms and Stochastic Approximations
TLDR
The juxtaposition of these two expressions in the title reflects the ambition of the authors to produce a reference work, both for engineers who use adaptive algorithms and for probabilists or statisticians who would like to study stochastic approximations in terms of problems arising from real applications.
Asymptotic Properties of Stochastic Approximations with Constant Coefficients.
Asymptotic properties (as $n \to \infty $, and then $a \to 0$) of the Stochastic Approximation (SA) algorithm \[ ( * )\qquad ,X_{n + 1}^a = X_n^a + ah_a \left( {X_n^a ,\xi _n^a } \right) \] are
Almost sure convergence