Optimization of mechanical properties of chitosan/phenol formaldehyde composite

Abstract

The Frechet distribution which has a scale and shape parameters, has been found to have wide application in modelling extreme events such as radioactive emission, flood, rainfall, seismic analysis, wind speed, etc. In this research paper, the Bayesian analysis of scale parameter of Frechet distribution was considered. It is necessary to know the best combination of prior distribution and loss function for the parameter estimation. Posterior distribution was derived by uniform and Jeffrey’s prior under the square error, Precautionary, Quadratic and Weighted balance loss function. Bayes estimation and their corresponding risk was obtained by the above stated priors and loss function. Monte Carlo simulations was conducted to compare the performance of the estimators. It is evident that weighted balance loss function when used with uniform prior provides the least posterior risk.

Keywords: Frechet Distribution, Non-Informative Prior, Bayesian Estimation, Loss Functions, Monte Carlo Simulations