作者alianlin (埋兵伏將)
看板Statistics
標題[問題] Indicator function
時間Tue Mar 28 22:58:17 2006
假如Y1.....Yn是random sample from uniform(a,b)
所以f(yi)=1/(b-a) , a<yi<b , i=1,2,...n
0 , 其他
又f(yi)=1/(b-a)可以表示成1/(b-a)‧I_(a,b) (yi) [_為下標]
n
那麼P(y1,y2......yn;a,b)=[1/(b-a)]^n‧Π I_(a,b) (yi) [因為彼此獨立]
i=1
=[1/(b-a)]^n‧ I_(a,y_(n)) (y_(1))‧I_(y_(1),b) (y_(n))
y_(n)=Max(Y1.....Yn) ,y_(1)=min(Y1.....Yn)
我想請問的是[1/(b-a)]^n‧ I_(a,y_(n)) (y_(1))‧I_(y_(1),b) (y_(n))這原式
是否也等同於[1/(b-a)]^n‧ I_(a,y_(n)) (y_(1))‧I_(a,b) (y_(n))呢?
這邊有點搞不懂,謝謝各位回答 ^^"
--
There is hope and consolation in the sure knowledge that
even the darkest hours of pains and troubles won't last.
--
※ 發信站: 批踢踢實業坊(ptt.cc)
◆ From: 140.112.215.175
※ 編輯: alianlin 來自: 140.112.215.175 (03/28 23:02)