[機統] continuous expectations and variances
定義出自 Saeed Ghahramani的機率課本 (閃電本) p.249
Theorem 6.2
For any continuous random variable X with probability distribution function
F and density function f,
∞ ∞
E(X) = ∫ [1 - F(t)] dt - ∫ F(-t) dt.
0 0
Proof: Note that
∞ 0 ∞
E(X) = ∫ xf(x) dx = ∫ xf(x) dx + ∫ xf(x) dx
-∞ -∞ 0
0 -x ∞ x
= -∫ (∫ dt)f(x) dx + ∫ (∫ dt)f(x) dx equation(1)
-∞ 0 0 0
∞ -t ∞ ∞
= -∫ (∫ f(x)dx) dt + ∫ (∫ f(x)dx) dt equation(2)
0 -∞ 0 t
where the last equality is obtained by changing the order of integration.
The theorem follows since
-t ∞
∫ f(x) dx = F(-t) and ∫f(x) dx = P(X > t) = 1 - F(t). (Proved!)
-∞ t
最後一行得證,是由 density function 的properties積分變成distribution的定義
去得到,這樣理解不知道有沒有錯?
主要想請問的是黃色的公式
-x x
Q1 期望值的xf(x)為什麼兩個x要換成 ∫ dt, ∫ dt 呢? 這個∫ dt 代表的意義是?
0 0
Q2 equation (1) 是怎麼轉換成 equation (2)的呢?
Q3 這邊的 積分範圍 & t 所代表的分別是什麼呢?
(自己的理解是,t為random variable X中某一點的位置,不過整個公式一起看
覺得自己並無法將整個公式的意義解釋清楚)
感謝
--
※ 發信站: 批踢踢實業坊(ptt.cc)
◆ From: 140.113.68.146
推
06/12 10:30, , 1F
06/12 10:30, 1F
→
06/12 10:30, , 2F
06/12 10:30, 2F
※ 編輯: Justin258 來自: 140.113.68.146 (06/12 12:25)
→
06/12 12:26, , 3F
06/12 12:26, 3F
→
06/12 12:27, , 4F
06/12 12:27, 4F
→
06/12 13:46, , 5F
06/12 13:46, 5F
→
06/12 13:56, , 6F
06/12 13:56, 6F
→
06/12 13:58, , 7F
06/12 13:58, 7F
→
06/12 13:58, , 8F
06/12 13:58, 8F
→
06/12 13:59, , 9F
06/12 13:59, 9F
→
06/12 15:11, , 10F
06/12 15:11, 10F
→
06/12 15:12, , 11F
06/12 15:12, 11F
→
06/14 00:12, , 12F
06/14 00:12, 12F
→
06/14 00:12, , 13F
06/14 00:12, 13F
→
06/14 00:13, , 14F
06/14 00:13, 14F
→
06/14 00:14, , 15F
06/14 00:14, 15F
→
06/14 00:14, , 16F
06/14 00:14, 16F