completeness

看板Statistics作者 (4/27視小中友之夜)時間19年前 (2006/06/03 22:10), 編輯推噓2(201)
留言3則, 2人參與, 最新討論串1/1
Def. Let f(s│θ) be a family of pdfs or pmfs for statistic S = S(X_1,...X_n). The family of probability distribution is called complete if E[u(S)]=0 for all θ implies u(S)=0 with probability 1 for all θ. Equivalently S is called a complete statistics. 上色的那句話 我看不懂是指什麼意思 可以請各位告訴我一下是什麼意思嗎? 謝謝 -- ※ 發信站: 批踢踢實業坊(ptt.cc) ◆ From: 140.112.243.58

06/04 02:16, , 1F
假設E[u(S)]=0,會得到 P(u(S)=0)=1
06/04 02:16, 1F

06/04 10:20, , 2F
E[u(S)]=0 => u(S)=0
06/04 10:20, 2F

06/04 10:23, , 3F
ie u(S)期望值等於零可推得其本身u(S)亦為零 即完備
06/04 10:23, 3F
文章代碼(AID): #14WPXcQH (Statistics)