Optimal adaptive filter realizations for sample stochastic processes with an unknown parameter
- 1 December 1969
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Automatic Control
- Vol. 14 (6), 767-770
- https://doi.org/10.1109/tac.1969.1099328
Abstract
Techniques are given for realizing optimal learning systems for filtering a sampled stochastic process in the presence of an unknown constant or time-varying parameter. It is shown how the nonlinear Bayes optimal (quadratic sense) adaptive filters can be directly realized for continuous parameter spaces by real-time analog systems. Examples are given for both constant and time-varying unknown parameters.Keywords
This publication has 5 references indexed in Scilit:
- Optimal Estimation in the Presence of Unknown ParametersIEEE Transactions on Systems Science and Cybernetics, 1969
- The prediction error of stationary Gaussian time series of unknown covarianceIEEE Transactions on Information Theory, 1965
- The filtering of time series with unknown signal statisticsPublished by Institute of Electrical and Electronics Engineers (IEEE) ,1965
- OPTIMAL ADAPTIVE ESTIMATION OF SAMPLED STOCHASTIC PROCESSESPublished by Defense Technical Information Center (DTIC) ,1963
- Adaptive communication filteringIEEE Transactions on Information Theory, 1962