Bayesian Point Prediction for Rayleigh distribution when observations are censored to left and right

AUTHOR AND
AFFILIATION

VASTOSHPATI SHASTRI
Department of Statistics Govt. P.G. Arts & Science College, Ratlam (India)
DEEPENDRA S. PAL
Department of Statistics Govt. P.G. Arts & Science College, Ratlam (India)

KEYWORDS:

Rayleigh Model

Issue Date:

February 2018

Pages:

97-109

ISSN:

2319-8044 (Online) – 2231-346X (Print)

Source:

Vol.30 – No.2

PDF

Click Here Download PDF

DOI:

http://dx.doi.org/10.22147/jusps-A/300201

ABSTRACT:

In this paper we utilize Bayesian approach to obtain predictors of the future observation from Rayleigh distribution when observations are censored to left as well as to the right. Bayesian predictor is obtained using natural conjugate prior under asymmetric loss function. Bayesian predictor is also obtained under the squared error loss function. For each loss predictive risks are calculated. Lastly, predictors are compared for the smallest future ordered observation on the basis of 1000 randomly generated sample using Monte Carlo simulation technique.

Copy the following to cite this Article:

V. Shastri; D. S. Pal, “Bayesian Point Prediction for Rayleigh distribution when observations are censored to left and right”, Journal of Ultra Scientist of Physical Sciences, Volume 30, Issue 2, Page Number 97-109, 2018


Copy the following to cite this URL:

V. Shastri; D. S. Pal, “Bayesian Point Prediction for Rayleigh distribution when observations are censored to left and right”, Journal of Ultra Scientist of Physical Sciences, Volume 30, Issue 2, Page Number 97-109, 2018

Available from: http://www.ultrascientist.org/paper/1151/bayesian-point-prediction-for-rayleigh-distribution-when-observations-are-censored-to-left-and-right


In this paper we utilize Bayesian approach to obtain predictors of the future observation from Rayleigh distribution when observations are censored to left as well as to the right. Bayesian predictor is obtained using natural conjugate prior under asymmetric loss function. Bayesian predictor is also obtained under the squared error loss function. For each loss predictive risks are calculated. Lastly, predictors are compared for the smallest future ordered observation on the basis of 1000 randomly generated sample using Monte Carlo simulation technique.