Title :
Minimum description length vs. maximum likelihood in lossy data compression
Author :
Madiman, M. ; Harrison, M. ; Kontoyiannis, I.
Author_Institution :
Div. of Appl. Math., Brown Univ., Providence, RI, USA
fDate :
27 June-2 July 2004
Abstract :
This paper describes the minimum description length principle in maximum likelihood estimate(MLE) in lossy data compression. In the lossless case the problem of optimal compression is theoretically equivalent to finding a probability distributions and minimizes the code-lengths.
Keywords :
codes; data compression; maximum likelihood estimation; probability; MLE; code-length; lossy data compression; maximum likelihood estimate; minimum description length; probability distribution; Compression algorithms; Data compression; Distortion measurement; Length measurement; Loss measurement; Mathematics; Maximum likelihood estimation; Probability distribution; Rate-distortion; US Department of Agriculture;
Conference_Titel :
Information Theory, 2004. ISIT 2004. Proceedings. International Symposium on
Print_ISBN :
0-7803-8280-3
DOI :
10.1109/ISIT.2004.1365499
