[機統] 關於機率論T分配的問題!

看板Math作者 (秘密客)時間13年前 (2012/09/19 13:02), 編輯推噓2(203)
留言5則, 2人參與, 最新討論串1/1
(a) Suppose that Xn~ tn follows a t-distribution with n degrees of freedom. Let fn(x) be the density of X. Find lim n->∞ fn(x). What does this tell you about the distribution of Xn for large n? Hint: Stirling’s approximation implies that lim n->∞ Γ(n).√n/2π.(e/n)^n=1. You may this useful for computing the desired limit. (b) Suppose that X1;X2; ::: ~ uniform(0; 1) are independent,so that P(X1< =x1; :::;Xn< =xn) =P(X1 < = x1).....P(Xn < = xn). Let Vn = n[1 - max{X1; ::;Xn}. Compute lim n->∞ Pn(Vn < = x) for x > 0. What does this tell you about the distribution of Vn when n is large? -- -- ※ 發信站: 批踢踢實業坊(ptt.cc) ◆ From: 204.52.215.1

09/19 15:03, , 1F
T分配自由度趨近無限大會變成標準常態分配,最簡單的
09/19 15:03, 1F

09/19 15:03, , 2F
證明可以用slusky定理說明就好
09/19 15:03, 2F

09/20 00:36, , 3F
(a)給hint:Stirling方程,你用t分配的pdf for n->∞
09/20 00:36, 3F

09/20 00:37, , 4F
會跟N(0.1)的pdf一樣.Stirling幫你處理Gamma的極限
09/20 00:37, 4F

09/20 00:38, , 5F
(b)Vn converges to Exp(1) in distribution.基本題
09/20 00:38, 5F
文章代碼(AID): #1GML7g6u (Math)