该文给出了一些负相协随机变量的指数不等式.这些不等式改进了由Jabbari 和Azarnoosh[4]及Oliveira[7] 所得到的相应的结果. 利用这些不等式对协方差系数为几何下降情形, 获得了强大数律的收敛速度为 n-1/2(log log n)1/2(log n)2. 这个收敛速度 接近独立随机变量的重对数律的收敛速度, 而Jabbari 和Azarnoosh[4]在上述情形下得到的收敛速度仅仅为n-1/3(log n)5/3.
The authors show some exponential inequalities for negatively associated random variables. These inequalities improve the rresponding results which Jabbari and Azarnoosh (2007) and Oliveira (2005) got. As an application, the authors obtain the rate of convergence n-1/2(log log n)1/2(log n)2 for the case of geometrically decreasing covariances, which closes to the optimal achievable convergence rate for independent random variables under an iterated logarithm, while Jabbari and Azarnoosh (2007) only got n-1/3(log n)5/3 for the case mentioned above.