Confidence Regions Based on Likelihood Ratio in a Nonlinear Model

[[abstract]]This paper is concerned with the confidence regions based on likelihood ratio in a nonlinear model. The quartic approximation of the coverage probability of the confidence regions is derived for the two-parameter nonlinear model. It is found that for the exponential and Michaelis-Menten models, the coverage probability is robust when the models deviate from the linearity assumptions