Variational optimization of the estimator stability

MaRDI QIDQ1922452zbMATH OpenFDO

Authors A. M. Shurygin

Publication date 7 January 1997

Published in Automation and Remote Control (Search for Journal in Brave)

zbMATH Keywords

estimator stabilityoptimization estimator

Mathematics Subject Classification ID

Signal detection and filtering (aspects of stochastic processes) (60G35) Filtering in stochastic control theory (93E11)

Review text

A variational measure is introduced to evaluate the instability of the estimator of the distribution density parameter with respect to small density variations. The estimation quality is characterized by two indexes, effectiveness and stability. A compromise estimator with the least loss in effectiveness and stability is computed. Maximal stable, compromise, and radical estimators of the parameters of an exponential, double exponential (Laplace), and normal distribution are given as examples.

Recommendations

• scientific article; zbMATH DE number 1848074
• Dimensions of multivariate statistics
• Quantification of estimation instability and its application to threshold selection in extremes
• Estimator stability in geological applications
• scientific article; zbMATH DE number 1834006

Cited in

(9)

• Asymptotically stable tests with application to robust detection
• scientific article; zbMATH DE number 1848074 (Why is no real title available?)
• Stable Limit Theory for the Variance Targeting Estimator
• Extremal variance and its generalization
• Optimal estimators for threshold-based quality measures
• Estimator stability in geological applications
• Analytic approach to variance optimization under an \(\mathcal{l}_1\) constraint
• Robust estimation of a correlation coefficient for \(\varepsilon\)-contaminated bivariate normal distributions
• Redescending \(M\)-estimators

This page was built for publication: Variational optimization of the estimator stability

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1922452)