Conditional empirical processes defined by nonstationary absolutely regular sequences (Q1303861)

\textit{K.-I. Yoshihara} [Comput. Math. Appl. 19, No. 1, 149-158 (1990; Zbl 0701.60024)] proved the weak invariance of the conditional nearest neighbour regression function estimator called the conditional empirical process based on \(\phi\)-mixing and identically distributed observations. The authors extend this result for nonstationary and absolutely regular sequences of random variables. This generalization makes it possible to deduce the weak invariance for Markov processes, for which the initial measure is not necessarily invariant.