The Appropriateness of the Asymptotic Distribution in a Nonlinear Regression Model in Relation to Curvature
非线性回归中渐近分布的适用性与曲率
SUMMARY In a non-linear regression model, correction terms for the asymptotic distribution of the maximum likelihood estimator are found. They can all be evaluated by the parameter dependent part of Beale's (1960) measure of non-linearity and some of the terms can be evaluated by Bates and Watts (1980) parameter dependent curvature in specific directions. The theory is illustrated by examples.