二项散布中方差的计算(国外英文资料).doc

二项分布中方差的计算(国外英文资料) The two distributions of Chinas poor calculation Lets say that xi ~ B of n, p So lets think about E of xi times xi minus one while make Type = on That is, And then I have E times xi equals np, and I have to do that The last too The distribution of case 1 The distribution of case 2 4.2 hypergeometric distribution Case 1 graphics: Case 2 graphics: Definition 4.2 there are two types of N elements, N1 belongs to the first category, N2 belong to the second (N1 + N2 = N). To take N sampling without replacement, from the factor said that N the number of the first kind of element, and the factor of distribution is called the hypergeometric distribution, Rule: if n r, then It is known by the nature of the probability distribution The properties of a combination There are two ways to calculate the mathematical expectation and variance of xi First, by definition K = m minus 1 Type = on This is the probability of only taking one of the elements N1 So the sampling (binomial distribution) is the same as the mathematical expectation of the sampling (hypergeometric distribution). For k = m - 2, Type = on so The q = 1 - p Another way of calculation is deduced mathematical expectation, assuming that factor I is the first time I had to draw the first kind of the number of elements, because only 1 time, of course not 1 is 0, so obey the 0-1 distribution, and there are , the so The whole method with the binomial distribution when the corresponding method is same, just the factor between the I had is not independent, but and expectation is equal to the desired and the properties are not required and the random variables are independent of each other. When N is very large, its much larger than the sampling N, and its N N super geometric distribution can be approximated by binomial distribution. To illustrate this point, the first approximation is the following: When N , N, we have This is because when N is large, the value behind each bracket is near to 1, so the a