zbMATH — the first resource for mathematics

Nonparametric signal estimation. (Neparametricheskoe otsenivanie signalov.) (Russian. English summary) Zbl 0886.62040
Moskva: Nauka. Fizmatlit. 336 p. (1997).
An approach is proposed for the analysis and synthesis of nonparametric kernel estimates for most practically applicable characteristics of conditional and unconditional distributions. The book presents methods of nonparametric statistics applicable to stationary dependent observations. In the first part, improved methods for the estimation of characteristics of a sequence of random variables are considered and their convergence in different metrics is studied. The second part is devoted to applications in identification of control systems and signal processing, with emphasis on the situation of a priori nonparametric uncertainty. The main authors’ idea is the unified functional approach, applied to a class of conditional distributions, as well as to decision procedures as functions of observations.
The book is addressed to specialists in information, control, filtering and extrapolation, who need estimation in the situation of essential a priori uncertainty.

62G07 Density estimation
62-02 Research exposition (monographs, survey articles) pertaining to statistics
94A12 Signal theory (characterization, reconstruction, filtering, etc.)
62P99 Applications of statistics
62M99 Inference from stochastic processes
62N99 Survival analysis and censored data
94-02 Research exposition (monographs, survey articles) pertaining to information and communication theory