【初见】蒙特卡洛采样
问题：理解蒙特卡洛采样公式
使用场景
数值求解积分，如：
y
=
x
2
,
x
∈
(
0
,
2
)
y=x^2,x\in (0,2)
y=x2,x∈(0,2)
-
均匀采样
I = b − a N ∑ i = 1 N f ( x i ) I = \frac{b-a}{N}\sum_{i=1}^{N}f(x_i) I=Nb−ai=1∑Nf(xi) -
非均匀采样
I = 1 N ∑ i = 1 N f ( x i ) p ( x i ) I = \frac{1}{N}\sum_{i=1}^{N}\frac{f(x_i)}{p(x_i)} I=N1i=1∑Np(xi)f(xi)
证明无偏估计
E
(
I
)
=
1
N
E
⟮
∑
i
=
1
N
f
(
x
i
)
p
(
x
i
)
⟯
=
1
N
∑
i
=
1
N
E
⟮
f
(
x
i
)
p
(
x
i
)
⟯
=
1
N
∑
i
=
1
N
∫
a
b
⟮
f
(
x
)
p
(
x
)
⟯
⋅
p
(
x
)
d
x
=
1
N
∑
i
=
1
N
∫
a
b
f
(
x
)
d
x
=
E
⟮
∫
a
b
f
(
x
)
d
x
⟯
=
∫
a
b
f
(
x
)
d
x
=
I
E(I)=\frac{1}{N}E\lgroup \sum_{i=1}^{N}\frac{f(x_i)}{p(x_i)}\rgroup \\ =\frac{1}{N}\sum_{i=1}^{N}E \lgroup{\frac{f(x_i)}{p(x_i)}}\rgroup\\=\frac{1}{N}\sum_{i=1}^{N}\int_{a}^{b}\lgroup{\frac{f(x)}{p(x)}}\rgroup \cdot p(x)dx\\=\frac{1}{N}\sum_{i=1}^{N}\int_{a}^{b}f(x)dx\\=E\lgroup{\int_{a}^{b}f(x)dx}\rgroup\\=\int_{a}^{b}f(x)dx\\=I
E(I)=N1E⟮i=1∑Np(xi)f(xi)⟯=N1i=1∑NE⟮p(xi)f(xi)⟯=N1i=1∑N∫ab⟮p(x)f(x)⟯⋅p(x)dx=N1i=1∑N∫abf(x)dx=E⟮∫abf(x)dx⟯=∫abf(x)dx=I
扫一扫
2万+
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
