Optimizing Optimization: The Next Generation of Optimization Applications and Theory (Quantitative Finance)

(Romina) #1

242 Optimizing Optimization


with


∂ ∂ ∂ ∂ ∂ ∂
L
i
L

L
i

ω

ωθμθ

θ

μω π

θ

ω



 

  

Ω 12

1

2

0

0

10

Consequently,


ωμ

θ
θ

Ω^11
2

(,)i









where


μμμ
μ

θ
θ

π





ΩΩ
ΩΩ

11
11

1
2 1

i
iii



























That is,


ωμˆ ˆ (ˆ,)ˆ

π
Ω^11
1

iQ









and therefore,


σωωπ

π

π

21

1

1
1

1

 

 





ˆ ˆˆ ()ˆ

ˆ ,

Ω Q

pQ p p

















where

From our earlier results, pQ p W T NTpQ p
ˆ^1 |ˆ ( , )
1
μ∼ 1 ,^11 where
W 1 ( · ) is a Wishart of dimension 1. Here, Qi i
(,)μμˆˆΩ^1 (),,
and therefore,
pQ p ˆ^1 ∼χ()^2 TN 1 Ψ, where χm


2
is a chi-squared with m degree of freedom.
with

Ψ

(^11)
T
pQ p

Free download pdf