Confidence intervals for probability density functions under strong mixing samples
Qingzhu Lei, Yongsong Qin · Journal of nonparametric statistics · 2015
It is shown that the empirical likelihood (EL) ratio statistic for a probability density function (p.d.f.) is asymptotically -type distributed under a strong mixing sample, which is used to obtain an EL-based confidence interval (CI) for the p.d.f. Results of a simulation study on the finite sample performance of the CI are reported.