Log-Exponentiated Frechet regression model with censored data

In this article, we propose the log-exponentiated Frechet regression model based on exponentiated Frechet distribution. Maximum likelihood and jackknife methods are used to estimate the Proposed model parameters based on censored data. Model assumptions are checked using martingale and deviance residuals. Moreover, global influence is used to detect influential observations. The behavior of the maximum likelihood and jackknife estimators is investigated through Monte Carlo simulation. The empirical distribution of the martingale and deviance residuals of the proposed model is examined. Leukemia data is analyzed to show the flexibility of the proposed model. (C) 2016 IASE Publisher. All rights reserved.