On the global central limit theorem for -mixing random variables

From MaRDI portal
Publication:1925167





Let \(\{X_j:j\geq 1\}\) be a \(\varphi\)-mixing sequence of r.v.'s such that \(EX^2_j<\infty\), \(EX_j=0\) and define \(Z_n= B_n^{-1}\sum^n_{j=1} X_j\) where \(B^2_n= E[(\sum^n_{j=1} X_j)^2]\). Several estimates of \[ \int^\infty_{-\infty} |x|^l|P(Z_n< x)-\Phi(x)|dx \] and, consequently, of \(\bigl|E|Z_n|^l- \int^\infty_{-\infty} |x|^l\Phi(dx)\bigr|\) (\(\Phi\) is the standard normal distribution function) are proved under mixing rate assumptions. The paper uses the method of \textit{C. Stein} [in: Proc. 6th Berkeley Sympos. Math. Statist. Probab., Univ. Calif. 1970, 2, 583-602 (1972; Zbl 0278.60026)] and continues previous work done by the author [Lith. Math. J. 22, 171-189 (1982); translation from Litov. Mat. Sb. 22, No. 2, 171-188 (1982; Zbl 0495.60039)] and by \textit{C. Cealera} [Stud. Cercet. Mat. 44, No. 1, 13-35 (1992; Zbl 0756.60019)].











This page was built for publication: On the global central limit theorem for \(\varphi\)-mixing random variables

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1925167)