11.07.2014 Views

Multiattribute acceptance sampling plans - Library(ISI Kolkata ...

Multiattribute acceptance sampling plans - Library(ISI Kolkata ...

Multiattribute acceptance sampling plans - Library(ISI Kolkata ...

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

2.3.5 Situation 4<br />

Theorem (2.3.4) :<br />

Assume N γ 1 , γ , p, p' are given.<br />

, 2<br />

There exists a minimum regret A plan with a = a<br />

p p '/<br />

p<br />

' j −1 / p<br />

j − 1 <<br />

j<br />

j<br />

j −1<br />

j<br />

, if<br />

….(2.3.10)<br />

Proof:<br />

Suppose there does not exist an optimal A plan with a j-1 = a j , but the given condition is<br />

satisfied.<br />

Let us suppose (a 1 ,…,a j-1 , a j ,..,a r ) is an optimal A plan with a j-1 < a j .<br />

Let us use the notation<br />

( j )<br />

a<br />

[ 1 2 j −1<br />

j r 1 2 j −<br />

j−1<br />

1<br />

∆ PA ( a , a ,..., a , a ,..., a ; m , m ,..., m , m ,..., m )] =<br />

= PA ( a1 ,...., a<br />

j −1<br />

, a<br />

j<br />

,..., a<br />

r<br />

; m1<br />

,..., mr<br />

) − PA(<br />

a1<br />

,..., a<br />

j −1<br />

, a<br />

j −1<br />

, a<br />

j + 1<br />

,..., a<br />

r<br />

; m<br />

1 ,..., mr<br />

)<br />

a1<br />

= ∑ ∑ ... ∑ ∏ g ( xi<br />

, mi<br />

) ∑ g ( x j −1<br />

, m j −1<br />

) g ( a j − x(<br />

j −1<br />

) , m j ) .<br />

x1<br />

= 0<br />

a2<br />

− x(1)<br />

x2<br />

= 0<br />

a j−2<br />

− x(<br />

j−3)<br />

j −2<br />

x j−2<br />

= 0<br />

i=<br />

1<br />

a j−1<br />

− x(<br />

j−2 )<br />

x j−1<br />

= 0<br />

j<br />

r<br />

a ' j+<br />

1 a ' j+<br />

2 − x j+<br />

1 a ' r −(<br />

x j+<br />

1 + x j+<br />

2 + ,... + xr−<br />

1 ) r<br />

∑ ∑ .....<br />

∑ ∏ g ( xi<br />

, mi<br />

)<br />

x = 0 x = 0<br />

x = 0 i=<br />

j + 1<br />

j+<br />

1<br />

Where ,<br />

a ' = a − a , ∀i<br />

≥ j + 1.<br />

i<br />

i<br />

j+<br />

2<br />

j<br />

Since (a 1 ,…,a j ,…,a r ) must satisfy the optimal criteria,<br />

r<br />

( j )<br />

a j −1<br />

…(2.3.11)<br />

∆ RA, ≤ 0 where the notation<br />

represents the difference of the regret function value for the A plan as explained in the proofs<br />

of theorems 2.3.2 and 2.3.3. The condition is equivalent to,<br />

( γ<br />

2<br />

≤ 1<br />

/ γ Exp m′<br />

1<br />

). (-( -m)).<br />

S<br />

A<br />

S<br />

A<br />

i = j − 2<br />

∏ h(<br />

x m′<br />

i<br />

,<br />

i<br />

)<br />

i = 1<br />

a<br />

j<br />

i = j − 2<br />

∏ h(<br />

x<br />

i<br />

, m<br />

i<br />

)<br />

i = 1<br />

A<br />

j<br />

−1<br />

− x<br />

( j − 2)<br />

i = r<br />

∑ h(<br />

x<br />

∏ ′<br />

j −1<br />

, m'<br />

j −1<br />

) h(<br />

a<br />

j<br />

− x<br />

( j −1)<br />

, m'<br />

j<br />

) S<br />

B<br />

h(<br />

x<br />

i<br />

, m<br />

i<br />

)<br />

x<br />

j −1<br />

= 0<br />

i = j + 1<br />

−1<br />

− x<br />

( j − 2)<br />

i = r<br />

∑ h(<br />

x<br />

j −1101<br />

, m<br />

j −1<br />

) h(<br />

A<br />

j<br />

− x<br />

( j −1)<br />

, m<br />

j<br />

) S<br />

B ∏ h(<br />

x<br />

i<br />

, m<br />

i<br />

)<br />

x<br />

j −1<br />

= 0<br />

i = j + 1

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!