Evaluation of Overflow Probability of Bayes Code in Moderate Deviation Regime
Shota Saito, Toshiyasu Matsushima · IEICE Transactions on Fundamentals of Electronics Communications and Computer Sciences · 2017
This paper treats the problem of lossless fixed-to-variable length source coding under the criterion of the overflow probability. We investigate the behavior of the overflow probability of the Bayes code in the moderate deviation regime. Our result clarifies that the behavior of the overflow probability of the Bayes code is the same as that of the optimal non-universal code in the moderate deviation regime.