M值随机序列加权和的强稳定性

The strong stability of weighted sums for M-valued random sequences

利用样本相对熵作为任意离散型随机变量相对独立变量偏差的一种随机性变量 ,采用研究强极限定理的一种新方法——网微分法 。

In this paper,the notion sample relative entropy as the random measure of deviation between discrete variables and independent random variables is introduced,we get a strong stability theorem for the weighted sums of discrete random sequences by using a new method of net differentiability.