Loading...
Please wait, while we are loading the content...
Similar Documents
Stochastic automata-based estimators for adaptively compressing files with non-stationary distributions.
| Content Provider | CiteSeerX |
|---|---|
| Author | Rueda, Luis Oommen, B. John |
| Abstract | In this paper, we show that learning automata (LA) techniques, which have been useful in developing weak estimators, can be applied to data compression applications in which the data distributions are non-stationary. The adaptive coding scheme utilizes Stochastic Learningbased Weak Estimation (SLWE) techniques to adaptively update the probabilities of the source symbols, and this is done without resorting to either maximum likelihood, Bayesian or slidingwindow methods. We have incorporated our estimator in the adaptive Fano coding scheme, and an adaptive entropy-based scheme that "resembles" the well-known arithmetic coding. The empirical results obtained for both these adaptive methods are obtained on real-life files that possess a fair degree of non-stationarity. From these results we see that our schemes compress nearly 10% more than their respective adaptive methods which use the MLE-based estimates. |
| File Format | |
| Access Restriction | Open |
| Subject Keyword | Non-stationary Distribution Stochastic Automata-based Estimator Real-life File Well-known Arithmetic Coding Data Compression Application Adaptive Method Slidingwindow Method Adaptive Coding Scheme Utilizes Stochastic Source Symbol Fair Degree Weak Estimation Respective Adaptive Method Adaptive Fano Maximum Likelihood Empirical Result Weak Estimator Adaptive Entropy-based Scheme Mle-based Estimate Data Distribution |
| Content Type | Text |
