UNIFORM CONVERGENCE RATES OF NORMAL APPROXIMATION FOR m-DEPENDENT RANDOM FIELDS

Let Z be the set of all positive integers, p∈Z, Zp={z=（z1,…,zp）, zj∈Z}. For any z and w∈Zp, define d（z, w）=max{|zj-wj|, 1≤j≤p}; and for any subsets V1 and V2 in Zp, define d（V1, V2）=inf {d（z, w）:z∈V1.W∈V2}.