Joint conditional quantiles estimation for strictly stationary process

Abstract EN

In this paper, the kernel estimation of the conditional quintiles for a strictly stationary stochastic process satisfying the strong mixing condition, which was proposed by Berger (1997), is studied.

Under some mild conditions, the joint asymptotic normality of the kernel estimation of several conditional quintiles estimated at the same conditional point and the joint asymptotic normality of the kernel estimation of the same conditional quintile estimated at different conditional points are established.

The performance of the estimations is tested by an application for a real life data.