Title :
Asymptotically minimax regret by Bayes mixtures
Author :
Takeuchi, Jun-ichi ; Barron, Andrew R.
Author_Institution :
C&C Media Res. Lab., NEC, Kanagawa, Japan
Abstract :
We study the problem of data compression, gambling and prediction of a sequence xn = x1x2...xn from a certain alphabet X, in terms of regret (Shtarkov 1988) and redundancy with respect to a general exponential family, a general smooth family, and also Markov sources. In particular, we show that variants of Jeffreys mixture asymptotically achieve their minimax values
Keywords :
Bayes methods; Markov processes; minimax techniques; prediction theory; redundancy; sequences; Bayes mixtures; Jeffreys mixture; Markov sources; data compression; gambling; general exponential family; general smooth family; prediction; redundancy; sequence; symptotically minimax regret; Data compression; Maximum likelihood estimation; Minimax techniques; National electric code; Statistics; Stochastic processes; Tail; USA Councils;
Conference_Titel :
Information Theory, 1998. Proceedings. 1998 IEEE International Symposium on
Conference_Location :
Cambridge, MA
Print_ISBN :
0-7803-5000-6
DOI :
10.1109/ISIT.1998.708923
