Strong optimality of the normalized ML models as universal codes and information in data

IEEE Transactions on Information Theory

Volumn 47, Issue 5, 2001, Pages 1712-1717

  Rissanen, J ^a  

^a IBM ALMADEN RESEARCH CENTER   (United States)

Author keywords

MDL principle; Minmax bounds; Noise; Relative redundancy; Stochastic complexity; Useful information

Indexed keywords

CODES (SYMBOLS); CONSTRAINT THEORY; DATA REDUCTION; MATHEMATICAL MODELS; MAXIMUM LIKELIHOOD ESTIMATION; NORMAL DISTRIBUTION; PROBABILITY DENSITY FUNCTION;

NORMALIZED MAXIMUM-LIKELIHOOD (NML) DISTRIBUTIONS; STOCHASTIC COMPLEXITY;

INFORMATION ANALYSIS;

EID: 0035396479     PISSN: 00189448     EISSN: None     Source Type: Journal    
DOI: 10.1109/18.930912     Document Type: Article

References (16)

1. Statistical inference, Occam's razor and statistical mechanics on the space of probability distributions
2. The MDL principle in modeling and coding
3. Information-theoretic asymptotics of Bayes methods
4. Universal noiseless coding
5. Iterated logarithmic expansions of the pathwise code lengths for exponential families
6. A strong version of the redundancy-capacity theorem of universal coding
7. Universal coding, information, prediction, and estimation
8. Stochastic complexity and modeling
9. Fisher information and stochastic complexity
10. MDL denoising
11. Universal sequential coding of single messages
12. Robustly minimax codes for universal data compression