|
Problemy Peredachi Informatsii, 1992, Volume 28, Issue 1, Pages 75–88
(Mi ppi1338)
|
|
|
|
Methods of Signal Processing
Convergence of Stochastic-Approximation Procedures in the Case of a Regression Equation with Several Roots
V. A. Lazarev
Abstract:
It is shown that the Robbins–Monro stochastic-approximation procedure cannot converge to the root $\tilde{x}$ of the regression equation $R(x)=0$ if at least one eigenvalue of the matrix $\frac{\partial R}{\partial x}(\tilde{x})$ has a positive real part. A similar result is obtained for the Kiefer–Wolfowitz procedure.
Received: 19.03.1991
Citation:
V. A. Lazarev, “Convergence of Stochastic-Approximation Procedures in the Case of a Regression Equation with Several Roots”, Probl. Peredachi Inf., 28:1 (1992), 75–88; Problems Inform. Transmission, 28:1 (1992), 66–78
Linking options:
https://www.mathnet.ru/eng/ppi1338 https://www.mathnet.ru/eng/ppi/v28/i1/p75
|
Statistics & downloads: |
Abstract page: | 319 | Full-text PDF : | 103 |
|