Comparison of Some Entropy Measures for Non-Central Fisher Probability Distribution

 

 

Mazeat Koreny1, Mohamed Bisher Zeina1, Shaza Zubeadah2

 

1 Department of Mathematical Statistics, Faculty of Science, University of Aleppo, Aleppo, Syria

2 Department of Basic Science, Faculty of Informatics Engineering, University of Aleppo, Aleppo, Syria

Emails: mazeat.koreny@gmail.com; bisher.zeina@gmail.com; szubeadah@gmail.com

*  Correspondence: bisher.zeina@gmail.com; Tel.: (+963969602951)

 

Abstract

In this paper, many entropy measures of noncentral Fisher distribution were driven including Shannon, Renyi, Sharma, Havrda, Arimoto and Tsallis. A comparison between these entropies was made according to distribution’s shift parameter, distribution’s degrees of freedom, shape parameter and truncation parameter. The entropy that had less relative loss was said to be better than the other. There were significant differences according to all studied parameters except the shift parameter and we found that the best entropy of the mentioned entropies for noncentral Fisher distribution was Renyi entropy.

Keywords: Goodness of Fit. Entropy; Noncentral Distributions; Relative Loss; Truncated Distribution; Shift Parameter.