Title :
On rate of convergence of some consistent estimates of the number of signals
Author :
Tam, E.W. ; Wu, Y.
Author_Institution :
Portland State Univ., OR, USA
Abstract :
Using the information theoretic criterion the authors previously obtained three consistent estimates of the number of signals for an additive model with white noise. In this paper the rates of convergence for the probabilities of wrong detections as a function of the sample size are studied. It is proved that under certain conditions and for a fairly general class of penalty terms, the probabilities of wrong detection are exponentially decreasing
Keywords :
convergence; estimation theory; information theory; probability; signal detection; white noise; additive model; consistent estimates; convergence rates; information theoretic criterion; number of signals; penalty terms; probabilities; white noise; Additive white noise; Convergence; Covariance matrix; Eigenvalues and eigenfunctions; Signal processing; State estimation; Statistical analysis; Statistics; Testing; White noise;
Conference_Titel :
Statistical Signal and Array Processing, 1996. Proceedings., 8th IEEE Signal Processing Workshop on (Cat. No.96TB10004
Conference_Location :
Corfu
Print_ISBN :
0-8186-7576-4
DOI :
10.1109/SSAP.1996.534816
