NEW CONCENTRATION INEQUALITIES AND COMPLETE CONVERGENCE FOR ELNQD RANDOM VARIABLES WITH APPLICATION TO LINEAR MODELS GENERATED BY ELNQD ERRORS

Abstract

In this paper, we introduce the concept of extended linear negative quadrant dependence (ELNQD, in short). We establish a new concentration inequalities and complete convergence for the distribution of sums of extended linear negative quadrant dependent random variables. Using these inequalities for proved the complete convergence of first autoregressive processes model generated by identically distributed ELNQD errors.