[機統] Covariance
Let X and Y be two identically distributed random variables.
Note that X and Y are not necessarily independent.
Prove
Cov(X + Y,X - Y ) = 0.
看到not necessarily independent就不知從何下手了
本來想用兩者獨立Cov為零的性質,顯然這題沒辦法。
觀念不是很清楚不知該怎麼處理QQ~
--
※ 發信站: 批踢踢實業坊(ptt.cc)
◆ From: 61.230.128.23
※ 編輯: spriteeddy 來自: 61.230.128.23 (06/17 16:40)
→
06/17 16:43, , 1F
06/17 16:43, 1F
→
06/17 16:44, , 2F
06/17 16:44, 2F
→
06/17 16:51, , 3F
06/17 16:51, 3F
→
06/17 16:51, , 4F
06/17 16:51, 4F
補問一題:
Suppose that X1;X2;X3 are independent Poisson random variables with respective
parameters λ1; λ2; λ3. Find Cov(X1 + X2;X2 + X3).
目前想法是把Cov(X1 + X2;X2 + X3)拆開
即Cov(X1 + X2;X2 + X3) = Cov(X1;X2) + Cov(X2;X2)
+ Cov(X1;X3) + Cov(X2;X3)
再用Covariance的定義把已知的值代進去
這樣想有問題嗎? 是否有其他想法??
※ 編輯: spriteeddy 來自: 61.230.128.23 (06/17 17:20)
→
06/17 18:18, , 5F
06/17 18:18, 5F
→
06/17 18:18, , 6F
06/17 18:18, 6F