摘要
For a linear model, let the error sequence be i.i.d, with common unknown density f(x), and (x) be a nonparametric estimator of f(x) based on the residuals. In this paper, on the basis of [1], we establish the L_1-norm consistency, asymptotic normality and law of iterated logarithm for (x) under general condition. These results bring the asymptotic theory for estimation of error distributions to completion.
For a linear model, let the error sequence be i.i.d, with common unknown density f（x）, and （x） be a nonparametric estimator of f（x） based on the residuals. In this paper, on the basis of [1], we establish the L<sub>1</sub>-norm consistency, asymptotic normality and law of iterated logarithm for （x） under general condition. These results bring the asymptotic theory for estimation of error distributions to completion.
基金
Project supported by the National Natural Science Foundation of China.
