多元正态分布的性质

在概率统计中,多元正态分布占有特别重要的地位.由于中心极限定理:独立随机变量序列之和的分布可以用正态分布来逼近、近似.而且,正态分布族密度函数只由其期望和协方差阵确定,不相关性与独立性一致,在建立概率模型时具有一定的优点。

The normal distribution occupies a very important po?sition. The main factors are based on random vector in the practical experience is often similar to the normal distribution and central lim?it theorems for the sum of independent distributed random vector. Therefore, characteristic function of multivariate normal distribution can be used to obtain first-order moments and second-order mo?ment in order to determine its distribution.