Title :
Exact minimax strategies for predictive density estimation, data compression and model selection
Author :
Liang, Feng ; Barron, Andrew R.
Author_Institution :
Inst. of Stat. & Decision Sci., Duke Univ., Durham, NC, USA
Abstract :
The Bayes procedure with uniform prior on location (and log-scale) parameters is shown to be exact minimax optimal for location and scale families in problems of universal data compression, predictive density estimation and model selection.
Keywords :
Bayes methods; data compression; information theory; minimax techniques; modelling; parameter estimation; prediction theory; Bayes procedure; exact minimax strategies; linear regression model; location parameters; model selection; predictive density estimation; universal data compression; Data compression; Linear regression; Minimax techniques; Predictive models; Q measurement; Redundancy; Statistical analysis; Statistical distributions; Statistics; Vectors;
Conference_Titel :
Information Theory, 2002. Proceedings. 2002 IEEE International Symposium on
Print_ISBN :
0-7803-7501-7
DOI :
10.1109/ISIT.2002.1023421
