(H0 : ✓ = 0) (H1 : ✓ = ✓1)
(H0 : ✓ = 0) (H1 : ✓ = ✓1)
↵
↵
Pr(P  ↵|H0) = ↵
Pr(P  ↵|H1) (1 )
R = Pr(H1)/Pr(Ho).
↵ R
↵%
H0 (1 ) H1
(✓ = ✓1)
(✓ = ✓1)
(✓ = 0)
(✓ = 0)
Pr(H0) = 1/(1 + R)
Pr(H1) = R/(1 +R).
Pr(H1\P  ↵)
Pr(H1\P  ↵)= Pr(H1)⇥Pr(P 
↵|H1)). Pr(P  ↵|H1)= (1 ) Pr(H1 \ P  ↵) = Pr(H1) (1 )
Pr(H1 \ P > ↵) Pr(H0 \ P > ↵)
Pr(H0 \ P < ↵)
Pr(P  ↵) = (R(1 ) + ↵) /(1 +R))
Pr(H1|P  ↵)
Pr(Ho|P  ↵)| {z } =
Pr(P  ↵|H1)
Pr(P  ↵|H0)| {z }⇥
Pr(H1)
Pr(Ho)| {z }
H1 H0
1
↵ ⇥R
Pr(H0|p < ↵)
(R,↵,) =
↵
R(1 ) + ↵
R(1 ) > ↵
90 R ⇡ 0.1
R = 0.1 ↵ = 0.05
R = 0.1 ↵ = 0.05
n 2 i = 1, ., n
Si = {gi : 0  gi  100}
gi
t? = 0 + p ⇥
1
n
P
gi p 2 (0, 1) 0 2 [0, 100).
di = |t? gi| t?
nwin di = dmin ⌘ min {d1, d2, ..., dn} .
ui(gi, gi) =
8<:$10/nwin, di dmin$0,
g? = {g?1 , g?2 , ..., g?n} = {0, 0, ..., 0}
0= 0
(n,0, p) =
(14, 20, 0.95) (14, 0, 0.95)
(n,0, p) = (14, 20, 0.95) (n,0, p) = (14, 0, 0.95)
t = 1, 2, ..., 10
x y