Abstract-- The problem of universal simulation given a training sequence is studied both in a stochastic setting and for individual sequences. In the stochastic setting, the traini...
We consider the problem of packetizing a variable length coded Markov sequence into fixed length packets, while being protected by variable rate channel code. Given the total tra...
— We propose an implementable new universal lossy source coding algorithm. The new algorithm utilizes two wellknown tools from statistical physics and computer science: Gibbs sam...
Background: In order to improve gene prediction, extrinsic evidence on the gene structure can be collected from various sources of information such as genome-genome comparisons an...
- The objective of this paper is to provide an effective technique for accurate modeling of the external input sequences that affect the behavior of Finite State Machines (FSMs). B...