Downloaded from https://www.everyspec.com V AMCP 706-197 " There is nothing mysterious about ran- the s-expected value of (x - p )n: donmess and random variables. If you need E{(x-)" Z (x,- 0)" pmf {x} something to be a random variable, it is; if (2-34) you don't need it to be, it isn't. where u E {x 1. 2-5.2 MOMENTS ' 2-5.3 TWO DISTRIBUTIONS Random variables with pmf's have mo- ments. Th2 two conventional points about Iwo common di,.crete distributions are which to take moments are the origin and the the binomial and Poisson. Table 21 gives mean; when taken about the mean, they are their definitions and properties. The Poisson called central-moments.
The second moment- distribution is often used as an approximation about-the-mean is the variance (square of the for the binomial distribution; it is usually ade- standard deviation). the s-expected value of xn; is negligible. nN+ The nth moment, about the origin, of x is quate if the Poisson probability pmf{n) E{xn }= 1;xpmf {x, 1 (2-33) where ' implies the sum over the domain of . i The adaptation (in most places) can be , x,. (It is presumed that the series converges made mechanically as follows: absolutely; if not, a textbook ought to be consulted.) PIN - p k~ - " oo The nth moment, about the mean, of x is P 2 1 2 2-11