A0676
Title: Maximum profile binomial likelihood estimation for the semiparametric Box--Cox power transformation model
Authors: Tao Yu - National University of Singapore (Singapore) [presenting]
Pengfei Li - University of Waterloo (Canada)
Baojiang Chen - University of Texas (United States)
Jing Qin - National Institutes of Health (United States)
Abstract: The Box-Cox transformation model has been widely applied for many years. The parametric version of this model assumes that the random error follows a parametric distribution, say the normal distribution, and estimates the model parameters using the maximum likelihood method. The semiparametric version assumes that the random error distribution is completely unknown; existing methods either need strong assumptions or are less effective when the random error distribution significantly deviates from the normal distribution. The semiparametric assumption is adopted, and a maximum profile binomial likelihood method is proposed. The joint distribution of the estimators of the model parameters is theoretically established. Through extensive numerical studies, it is demonstrated that this method has an advantage over existing methods when the distribution of the random error deviates from the normal distribution. Furthermore, the method's performance is compared to existing methods on an HIV data set.
