Abstract
An adaptive, universal VF lossless source coding technique that generates a dictionary sequentially adapted to the empirical distribution of the observed data is presented in this study. In view of the variations of the dictionary, an algorithm is proposed for implementing the VF code. For i.i.d. sources the proposed algorithm converges to a specific entropy. Finally, the adaptive dictionary construction algorithm has some resemblance to arithmetic coding.
Original language | English |
---|---|
State | Published - 1994 |
Event | Proceedings of the 1994 IEEE International Symposium on Information Theory - Trodheim, Norw Duration: 27 Jun 1994 → 1 Jul 1994 |
Conference
Conference | Proceedings of the 1994 IEEE International Symposium on Information Theory |
---|---|
City | Trodheim, Norw |
Period | 27/06/94 → 1/07/94 |