概率论与数理统计期末总结其二
博主是21级计科院的,概率论与数理统计95分,这是期末复习期间整理的笔记(其二),基本全部涵盖期末考试重点范围,有需要的学弟学妹可以作为参考
第二讲 方差
2.1 方差
D
X
=
E
(
X
−
E
X
)
2
⇒
D
X
=
E
X
2
−
(
E
X
)
2
(
推论,可以转求
E
X
2
)
DX = E(X-EX)^2 \Rightarrow DX = EX^2-(EX)^2(推论,可以转求EX^2)
DX=E(X−EX)2⇒DX=EX2−(EX)2(推论,可以转求EX2)
其中
D
X
\sqrt{DX}
DX 为随机变量
X
X
X 的均方差或标准差;
2.2 性质
D
(
C
1
X
+
C
2
)
=
C
1
2
D
X
D
(
X
±
Y
)
=
D
X
+
D
Y
±
2
c
o
v
(
X
,
Y
)
D
X
=
0
⟺
P
(
X
=
E
X
)
=
1
D(C_1X+C_2)=C_1^2DX\\ D(X\pm Y)=DX+DY\pm 2cov(X,Y)\\ DX=0\Longleftrightarrow P(X=EX)=1
D(C1X+C2)=C12DXD(X±Y)=DX+DY±2cov(X,Y)DX=0⟺P(X=EX)=1
【
T
h
e
o
r
e
m
Theorem
Theorem】
X
i
∼
N
(
μ
i
,
σ
i
2
)
(
i
=
1
,
2
,
⋅
⋅
⋅
,
n
)
X_i\thicksim N(\mu_i,\sigma_i^2)(i=1,2,\cdot\cdot\cdot,n)
Xi∼N(μi,σi2)(i=1,2,⋅⋅⋅,n) 相互独立,则
c
1
X
1
+
c
2
X
2
+
⋅
⋅
⋅
+
c
n
X
n
∼
N
(
c
1
μ
1
+
c
2
μ
2
+
⋅
⋅
⋅
+
c
n
μ
n
,
c
1
2
σ
1
2
+
c
2
2
σ
2
2
+
⋅
⋅
⋅
+
c
n
2
σ
n
2
)
(
c
1
2
+
c
2
2
+
⋅
⋅
⋅
+
c
n
2
≠
0
)
c_1X_1+c_2X_2+\cdot\cdot\cdot+c_nX_n\thicksim N(c_1\mu_1+c_2\mu_2+\cdot\cdot\cdot+c_n\mu_n,c_1^2\sigma_1^2+c_2^2\sigma_2^2+\cdot\cdot\cdot+c_n^2\sigma_n^2)\\(c_1^2+c_2^2+\cdot\cdot\cdot+c_n^2\neq0)
c1X1+c2X2+⋅⋅⋅+cnXn∼N(c1μ1+c2μ2+⋅⋅⋅+cnμn,c12σ12+c22σ22+⋅⋅⋅+cn2σn2)(c12+c22+⋅⋅⋅+cn2=0)
第三讲 协方差与相关系数
3.1 协方差
c o v ( X , Y ) = E [ ( X − E X ) ( Y − E Y ) ] ⇒ c o v ( X , Y ) = E ( X Y ) − E X ⋅ E Y cov(X,Y)=E[(X-EX)(Y-EY)]\\ \Rightarrow cov(X,Y)=E(XY)-EX\cdot EY cov(X,Y)=E[(X−EX)(Y−EY)]⇒cov(X,Y)=E(XY)−EX⋅EY
3.2 性质
- c o v ( X , X ) = D X , c o v ( X , C ) = 0 cov(X,X)=DX,cov(X,C)=0 cov(X,X)=DX,cov(X,C)=0
- c o v ( X , Y ) = c o v ( Y , X ) cov(X,Y)=cov(Y,X) cov(X,Y)=cov(Y,X)
- c o v ( a X 1 + b X 2 , c Y + d ) = a c c o v ( X 1 , Y ) + b c c o v ( X 2 , Y ) cov(aX_1+bX_2,cY+d)=accov(X_1,Y)+bccov(X_2,Y) cov(aX1+bX2,cY+d)=accov(X1,Y)+bccov(X2,Y)
- D ( X ± Y ) = D X + D Y ± 2 c o v ( X , Y ) D(X\pm Y)=DX+DY\pm 2cov(X,Y) D(X±Y)=DX+DY±2cov(X,Y)
3.3 相关系数
ρ
X
Y
=
c
o
v
(
X
,
Y
)
D
X
D
Y
(
D
X
>
0
,
D
Y
>
0
)
\rho_{XY}=\frac{cov(X,Y)}{\sqrt{DX}\sqrt{DY}}(DX>0,DY>0)
ρXY=DXDYcov(X,Y)(DX>0,DY>0)
注解:转向求协方差
【
T
h
e
o
r
e
m
Theorem
Theorem】
(
X
,
Y
)
∼
N
(
μ
1
,
μ
2
;
σ
1
2
,
σ
2
2
;
ρ
)
⇒
ρ
X
Y
=
ρ
⇒
X
与
Y
相互独立
⟺
不相关
(X,Y)\thicksim N(\mu_1,\mu_2;\sigma_1^2,\sigma_2^2;\rho)\Rightarrow \rho_{XY}=\rho\\ \Rightarrow X与Y相互独立\Longleftrightarrow不相关
(X,Y)∼N(μ1,μ2;σ12,σ22;ρ)⇒ρXY=ρ⇒X与Y相互独立⟺不相关
3.4 C a u c h y − S c h w a r z Cauchy-Schwarz Cauchy−Schwarz 不等式
[ E ( X Y ) ] 2 ≤ E X 2 E Y 2 , ( E X 2 < + ∞ , E Y 2 < ∞ ) [E(XY)]^2\leq EX^2EY^2,(EX^2<+\infty,EY^2<\infty) [E(XY)]2≤EX2EY2,(EX2<+∞,EY2<∞)
3.5 性质
由柯西施瓦茨不等式推得: ∣ ρ X Y ∣ ≤ 1 |\rho_{XY}|\leq 1 ∣ρXY∣≤1
∣ ρ X Y ∣ = 1 ⟺ ∃ a ( a ≠ 0 ) , b ⇒ P ( Y = a X + b ) = 1 |\rho_{XY}| = 1 \Longleftrightarrow \exists a(a\neq0), b\Rightarrow P(Y=aX+b)=1 ∣ρXY∣=1⟺∃a(a=0),b⇒P(Y=aX+b)=1
3.6 不相关
D X > 0 , D Y > 0 DX>0,DY>0 DX>0,DY>0
- X , Y X,Y X,Y 不相关 ⟺ c o v ( X , Y ) = 0 \Longleftrightarrow cov(X,Y)=0 ⟺cov(X,Y)=0
- X , Y X,Y X,Y 不相关 ⟺ E ( X Y ) = E X ⋅ E Y \Longleftrightarrow E(XY)=EX\cdot EY ⟺E(XY)=EX⋅EY
- X , Y X,Y X,Y 不相关 ⟺ D ( X ± Y ) = D X + D Y \Longleftrightarrow D(X\pm Y) = DX+DY ⟺D(X±Y)=DX+DY
- X , Y X,Y X,Y 不相关 ⟺ D ( X + Y ) = D ( X − Y ) \Longleftrightarrow D(X+Y)=D(X-Y) ⟺D(X+Y)=D(X−Y)
第四讲 n n n 维正态随机变量
4.1 矩的概念
- k k k 阶原点矩 μ k = E X k , k = 1 , 2 , ⋅ ⋅ ⋅ \mu_k=EX^k,k = 1,2,\cdot\cdot\cdot μk=EXk,k=1,2,⋅⋅⋅
- k k k 阶中心矩 ν k = E ( X − E X ) k , k = 1 , 2 , ⋅ ⋅ ⋅ \nu_k=E(X-EX)^k,k = 1,2,\cdot\cdot\cdot νk=E(X−EX)k,k=1,2,⋅⋅⋅
- k + l k+l k+l 阶混合原点矩 μ k l = E ( X k Y l ) , k , l = 1 , 2 , ⋅ ⋅ ⋅ \mu_{kl}=E(X^kY^l),k,l = 1,2,\cdot\cdot\cdot μkl=E(XkYl),k,l=1,2,⋅⋅⋅
- k + l k+l k+l 阶混合中心矩 ν k l = E [ ( X − E X ) k ( Y − E Y ) l ] , k , l = 1 , 2 , ⋅ ⋅ ⋅ \nu_{kl}=E[(X-EX)^k(Y-EY)^l],k,l = 1,2,\cdot\cdot\cdot νkl=E[(X−EX)k(Y−EY)l],k,l=1,2,⋅⋅⋅
4.2 均值向量
X = ( X 1 , X 2 , ⋅ ⋅ ⋅ , X n ) , μ i = E X i ( i = 1 , 2 , ⋅ ⋅ ⋅ , n ) ⇒ μ = ( μ 1 , μ 2 , ⋅ ⋅ ⋅ , μ n ) X=(X_1,X_2,\cdot\cdot\cdot,X_n),\mu_i=EX_i(i=1,2,\cdot\cdot\cdot,n)\\ \Rightarrow \boldsymbol{\mu}=(\mu_1,\mu_2,\cdot\cdot\cdot,\mu_n) X=(X1,X2,⋅⋅⋅,Xn),μi=EXi(i=1,2,⋅⋅⋅,n)⇒μ=(μ1,μ2,⋅⋅⋅,μn)
4.3 协方差矩阵
σ i j = c o v ( X i , X j ) = E [ ( X i − E X i ) ( X j − E X j ) ] ( i , j = 1 , 2 , ⋅ ⋅ ⋅ , n ) B = ( σ i j ) n × n = [ σ 11 σ 12 ⋯ σ 1 n σ 21 σ 22 ⋯ σ 2 n ⋮ ⋮ ⋯ ⋮ σ n 1 σ n 2 ⋯ σ n n ] \sigma_{ij}=cov(X_i,X_j)=E[(X_i-EX_i)(X_j-EX_j)](i,j=1,2,\cdot\cdot\cdot,n)\\ \boldsymbol{B}=(\sigma_{ij})_{n\times n}= \left[\begin{array}{cccc} \sigma_{11} & \sigma_{12} & \cdots & \sigma_{1n}\\ \sigma_{21} & \sigma_{22} & \cdots & \sigma_{2n}\\ \vdots & \vdots & \cdots & \vdots\\ \sigma_{n1} & \sigma_{n2} & \cdots & \sigma_{nn} \end{array}\right] σij=cov(Xi,Xj)=E[(Xi−EXi)(Xj−EXj)](i,j=1,2,⋅⋅⋅,n)B=(σij)n×n= σ11σ21⋮σn1σ12σ22⋮σn2⋯⋯⋯⋯σ1nσ2n⋮σnn
4.4 n n n 维正态随机变量
X = ( X 1 , X 2 , ⋯ , X n ) , μ = ( μ 1 , μ 2 , ⋅ ⋅ ⋅ , μ n ) , B = ( σ i j ) n × n 若其联合概率密度 f ( x ) = f ( x 1 , x 2 , ⋯ , x n ) = 1 ( 2 π ) n 2 ∣ B ∣ 1 2 e − 1 2 ( x − μ ) B − 1 ( x − μ ) T ( i = 1 , 2 , ⋯ , n ) ⇒ X = ( X 1 , X 2 , ⋯ , X n ) ∼ N ( μ , B ) \boldsymbol{X}=(X_1,X_2,\cdots,X_n),\boldsymbol{\mu}=(\mu_1,\mu_2,\cdot\cdot\cdot,\mu_n),\boldsymbol{B}=(\sigma_{ij})_{n\times n}\\ 若其联合概率密度 f(\boldsymbol{x})=f(x_1,x_2,\cdots,x_n)=\frac{1}{(2\pi)^\frac{n}{2}|\boldsymbol{B}|^\frac{1}{2}}e^{-\frac{1}{2}(\boldsymbol{x}-\boldsymbol{\mu})\boldsymbol{B}^{-1}(\boldsymbol{x}-\boldsymbol{\mu})^T}(i=1,2,\cdots,n)\\ \Rightarrow \boldsymbol{X}=(X_1,X_2,\cdots,X_n)\thicksim N(\boldsymbol{\mu},\boldsymbol{B}) X=(X1,X2,⋯,Xn),μ=(μ1,μ2,⋅⋅⋅,μn),B=(σij)n×n若其联合概率密度f(x)=f(x1,x2,⋯,xn)=(2π)2n∣B∣211e−21(x−μ)B−1(x−μ)T(i=1,2,⋯,n)⇒X=(X1,X2,⋯,Xn)∼N(μ,B)
4.5 性质
- { X ∼ N ( μ , B ) ⇒ X i ∼ N ( μ i , σ i i ) ( i = 1 , 2 , ⋯ , n ) X i ∼ N ( μ i , σ i i ) ( i = 1 , 2 , ⋯ , n ) 且相互独立 ⇒ X ∼ N ( μ , B ) \left\{\begin{array}{l}\boldsymbol{X}\thicksim N(\boldsymbol{\mu},\boldsymbol{B})\Rightarrow X_i\thicksim N(\mu_i,\sigma_{ii})(i=1,2,\cdots,n)\\X_i\thicksim N(\mu_i,\sigma_{ii})(i=1,2,\cdots,n) 且相互独立\Rightarrow \boldsymbol{X}\thicksim N(\boldsymbol{\mu},\boldsymbol{B})\end{array}\right. {X∼N(μ,B)⇒Xi∼N(μi,σii)(i=1,2,⋯,n)Xi∼N(μi,σii)(i=1,2,⋯,n)且相互独立⇒X∼N(μ,B)
- X ∼ N ( μ , B ) ⟺ ∀ c 1 , c 2 , ⋯ , c n , 满足 c 1 2 + c 2 2 + ⋅ ⋅ ⋅ + c n 2 ≠ 0 ,有 c 1 X 1 + c 2 X 2 + ⋯ + c n X n ∼ N ( μ , σ ) \boldsymbol{X}\thicksim N(\boldsymbol{\mu},\boldsymbol{B})\Longleftrightarrow \forall c_1,c_2,\cdots,c_n,满足c_1^2+c_2^2+\cdot\cdot\cdot+c_n^2\neq0,有c_1X_1+c_2X_2+\cdots+c_nX_n\thicksim N(\mu,\sigma) X∼N(μ,B)⟺∀c1,c2,⋯,cn,满足c12+c22+⋅⋅⋅+cn2=0,有c1X1+c2X2+⋯+cnXn∼N(μ,σ)
- X ∼ N ( μ , B ) , ( Y 1 , Y 2 , ⋯ , Y k ) 满足 Y i = α T X ( i = 1 , 2 , ⋯ , k ) , 则 Y = ( Y 1 , Y 2 , ⋯ , Y k ) ∼ N ( μ , σ ) \boldsymbol{X}\thicksim N(\boldsymbol{\mu},\boldsymbol{B}),(Y_1,Y_2,\cdots,Y_k)满足Y_i=\boldsymbol{\alpha}^T\boldsymbol{X}(i=1,2,\cdots,k),则\boldsymbol{Y}=(Y_1,Y_2,\cdots,Y_k)\thicksim N(\mu,\sigma) X∼N(μ,B),(Y1,Y2,⋯,Yk)满足Yi=αTX(i=1,2,⋯,k),则Y=(Y1,Y2,⋯,Yk)∼N(μ,σ)
- X = ( X 1 , X 2 , ⋯ , X n ) ∼ N ( μ , B ) ⇒ X ′ = ( X 1 , X 2 , ⋯ , X m ) ∼ N ( μ ′ , B ′ ) ( m < n ) \boldsymbol{X}=(X_1,X_2,\cdots,X_n)\thicksim N(\boldsymbol{\mu},\boldsymbol{B})\Rightarrow \boldsymbol{X}'=(X_1,X_2,\cdots,X_m)\thicksim N(\boldsymbol{\mu}',\boldsymbol{B}')(m<n) X=(X1,X2,⋯,Xn)∼N(μ,B)⇒X′=(X1,X2,⋯,Xm)∼N(μ′,B′)(m<n)
第五章 大数定理及中心极限定理
第一讲 切比雪夫不等式
{ P ( ∣ X − E X ∣ ≥ ε ) ≤ D X ε 2 P ( ∣ X − E X ∣ < ε ) ≥ 1 − D X ε 2 \left\{\begin{array}{c}P(|X-EX|\geq \varepsilon)\leq \frac{DX}{\varepsilon^2}\\P(|X-EX|<\varepsilon)\geq 1-\frac{DX}{\varepsilon^2}\end{array}\right. {P(∣X−EX∣≥ε)≤ε2DXP(∣X−EX∣<ε)≥1−ε2DX
第二讲 大数定律
随机变量序列的前若干项的算术平均值在某种条件下收敛到这些项的均值的算术平均值
| lim n → ∞ P ( ∣ 1 n ∑ i = 1 n X i − a n ∣ ≥ ε ) = 0 \lim_{n\rightarrow \infty}P(|\frac{1}{n}\sum_{i=1}^nX_i-a_n|\geq \varepsilon)=0 limn→∞P(∣n1∑i=1nXi−an∣≥ε)=0 | 随机变量前 n n n 项均值依概率收敛于某一定值 |
|:----😐:----😐
| lim n → ∞ P ( ∣ X n − X ∣ ≥ ε ) = 0 即 X n → P X ( n → ∞ ) \lim_{n\rightarrow \infty}P(|X_n-X|\geq \varepsilon)=0即X_n\rightarrow^PX(n\rightarrow\infty) limn→∞P(∣Xn−X∣≥ε)=0即Xn→PX(n→∞)|随机变量依概率收敛于另一随机变量|
【大数定理】
大数定理 条件 定律形式 切比雪夫大数定律 X i 相互独立 , E X i = μ = C 1 , D X i = σ 2 = C 2 X_i相互独立,EX_i=\mu=C_1,DX_i=\sigma^2=C_2 Xi相互独立,EXi=μ=C1,DXi=σ2=C2 $\begin{array}{c}lim_{n\rightarrow \infty}P( 切比雪夫大数定律一般形式 X i 相互独立 , E X i = μ = C 1 , D X i ≤ C X_i相互独立,EX_i=\mu=C_1,DX_i\leq C Xi相互独立,EXi=μ=C1,DXi≤C $\begin{array}{c}lim_{n\rightarrow \infty}P( 马尔可夫大数定理 lim n → ∞ 1 n 2 D [ ∑ i = 1 n X i ] = 0 \lim_{n\rightarrow \infty}\frac{1}{n^2}D[\sum_{i=1}^nX_i]=0 limn→∞n21D[∑i=1nXi]=0 $\begin{array}{c}\lim_{n\rightarrow \infty}P( 辛钦大数定理 X i 相互独立同分布 , E X i = μ = C < ∞ X_i相互独立同分布,EX_i=\mu=C<\infty Xi相互独立同分布,EXi=μ=C<∞ $\begin{array}{c}\lim_{n\rightarrow \infty}P( 伯努利大数定理 f A = n A n , P ( A ) = p f_A=\frac{n_A}{n},P(A)=p fA=nnA,P(A)=p $\begin{array}{c}lim_{n\rightarrow \infty}P(
第三讲 中心极限定理
3.1 独立同分布中心极限定理
【适用条件】 X i 相互独立且同分布 ⇒ E X i = μ , D X i = σ 2 ( i = 1 , 2 , ⋯ ) 令随机变量 Y n = ∑ i = 1 n X i − n μ n σ ,其分布函数 F n ( x ) lim n → ∞ F n ( x ) = lim n → ∞ P ( ∑ i = 1 n X i − n μ n σ ≤ x ) = ∫ − ∞ x 1 2 π e − t 2 2 d t = Φ ( x ) ⟹ Y i = ∑ i = 1 n X i ∼ N ( n μ , n σ 2 ) ( n → ∞ ) ⟹ { P ( ∑ i = 1 n X i ≤ x ) ≈ Φ ( x − n μ n σ ) P ( a < ∑ i = 1 n X i ≤ b ) ≈ Φ ( b − n μ n σ ) − Φ ( a − n μ n σ ) 【适用条件】X_i相互独立且同分布\Rightarrow EX_i=\mu,DX_i=\sigma^2(i=1,2,\cdots)\\ 令随机变量Y_n=\frac{\sum_{i=1}^{n}X_i-n\mu}{\sqrt{n}\sigma},其分布函数F_n(x)\\ \lim_{n\rightarrow \infty}F_n(x)=\lim_{n\rightarrow \infty}P(\frac{\sum_{i=1}^{n}X_i-n\mu}{\sqrt{n}\sigma}\leq x)=\int_{-\infty}^x\frac{1}{\sqrt{2\pi}}e^{-\frac{t^2}{2}}dt=\Phi(x)\\ \Longrightarrow Y_i=\sum_{i=1}^nX_i\thicksim N(n\mu,n\sigma^2)(n\rightarrow \infty)\\ \Longrightarrow \left\{\begin{array}{c}P(\sum_{i=1}^nX_i\leq x)\approx \Phi(\frac{x-n\mu}{\sqrt{n}\sigma})\\ P(a<\sum_{i=1}^nX_i\leq b)\approx \Phi(\frac{b-n\mu}{\sqrt{n}\sigma})-\Phi(\frac{a-n\mu}{\sqrt{n}\sigma})\end{array}\right. 【适用条件】Xi相互独立且同分布⇒EXi=μ,DXi=σ2(i=1,2,⋯)令随机变量Yn=nσ∑i=1nXi−nμ,其分布函数Fn(x)n→∞limFn(x)=n→∞limP(nσ∑i=1nXi−nμ≤x)=∫−∞x2π1e−2t2dt=Φ(x)⟹Yi=i=1∑nXi∼N(nμ,nσ2)(n→∞)⟹{P(∑i=1nXi≤x)≈Φ(nσx−nμ)P(a<∑i=1nXi≤b)≈Φ(nσb−nμ)−Φ(nσa−nμ)
3.2 李雅普诺夫中心极限定理
【适用条件】 X i 相互独立且 E X i = μ i , D X i = σ i 2 ( i = 1 , 2 , ⋯ ) 令随机变量 Y n = ∑ i = 1 n X i − E [ ∑ i = 1 n X i ] D [ ∑ i = 1 n X i ] = ∑ i = 1 n X i − ∑ i = 1 n μ i ∑ i = 1 n σ i 2 ,其分布函数 F n ( x ) lim n → ∞ F n ( x ) = lim n → ∞ P ( ∑ i = 1 n X i − ∑ i = 1 n μ i ∑ i = 1 n σ i 2 ≤ x ) = ∫ − ∞ x 1 2 π e − t 2 2 d t = Φ ( x ) ⟹ Y i = ∑ i = 1 n X i ∼ N ( ∑ i = 1 n μ i , ∑ i = 1 n σ i 2 ) ( n → ∞ ) ⟹ { P ( ∑ i = 1 n X i ≤ x ) ≈ Φ ( x − ∑ i = 1 n μ i ∑ i = 1 n σ i 2 ) P ( a < ∑ i = 1 n X i ≤ b ) ≈ Φ ( b − ∑ i = 1 n μ i ∑ i = 1 n σ i 2 ) − Φ ( a − ∑ i = 1 n μ i ∑ i = 1 n σ i 2 ) 【适用条件】X_i相互独立且EX_i=\mu_i,DX_i=\sigma_i^2(i=1,2,\cdots)\\ 令随机变量Y_n=\frac{\sum_{i=1}^{n}X_i-E[\sum_{i=1}^{n}X_i]}{\sqrt{D[\sum_{i=1}^nX_i]}}=\frac{\sum_{i=1}^nX_i-\sum_{i=1}^n\mu_i}{\sqrt{\sum_{i=1}^n\sigma_i^2}},其分布函数F_n(x)\\ \lim_{n\rightarrow \infty}F_n(x)=\lim_{n\rightarrow \infty}P(\frac{\sum_{i=1}^{n}X_i-\sum_{i=1}^n\mu_i}{\sqrt{\sum_{i=1}^n\sigma_i^2}}\leq x)=\int_{-\infty}^x\frac{1}{\sqrt{2\pi}}e^{-\frac{t^2}{2}}dt=\Phi(x)\\ \Longrightarrow Y_i=\sum_{i=1}^nX_i\thicksim N(\sum_{i=1}^n\mu_i,\sum_{i=1}^n\sigma_i^2)(n\rightarrow \infty)\\ \Longrightarrow \left\{\begin{array}{c}P(\sum_{i=1}^nX_i\leq x)\approx \Phi(\frac{x-\sum_{i=1}^n\mu_i}{\sqrt{\sum_{i=1}^n\sigma_i^2}})\\ P(a<\sum_{i=1}^nX_i\leq b)\approx \Phi(\frac{b-\sum_{i=1}^n\mu_i}{\sqrt{\sum_{i=1}^n\sigma_i^2}})-\Phi(\frac{a-\sum_{i=1}^n\mu_i}{\sqrt{\sum_{i=1}^n\sigma_i^2}})\end{array}\right. 【适用条件】Xi相互独立且EXi=μi,DXi=σi2(i=1,2,⋯)令随机变量Yn=D[∑i=1nXi]∑i=1nXi−E[∑i=1nXi]=∑i=1nσi2∑i=1nXi−∑i=1nμi,其分布函数Fn(x)n→∞limFn(x)=n→∞limP(∑i=1nσi2∑i=1nXi−∑i=1nμi≤x)=∫−∞x2π1e−2t2dt=Φ(x)⟹Yi=i=1∑nXi∼N(i=1∑nμi,i=1∑nσi2)(n→∞)⟹⎩ ⎨ ⎧P(∑i=1nXi≤x)≈Φ(∑i=1nσi2x−∑i=1nμi)P(a<∑i=1nXi≤b)≈Φ(∑i=1nσi2b−∑i=1nμi)−Φ(∑i=1nσi2a−∑i=1nμi)
注重两个中心极限定理的区别与联系
3.3 D e M o i v r e − L a p l a c e De Moivre-Laplace DeMoivre−Laplace 中心极限定理
【 n 重伯努利试验】 f A = n A n , P ( A ) = p 令随机变量 Y n = n A − n p n p ( 1 − p ) ,其分布函数 F n ( x ) lim n → ∞ F n ( x ) = lim n → ∞ P ( n A − n p n p ( 1 − p ) ≤ x ) = ∫ − ∞ x 1 2 π e − t 2 2 d t = Φ ( x ) ⟹ X ∼ N ( n p , n p ( 1 − p ) ) ( n → ∞ ) ⟹ { P ( X ≤ x ) ≈ Φ ( x − n p n p ( 1 − p ) ) P ( a < X ≤ b ) ≈ Φ ( b − n p n p ( 1 − p ) ) − Φ ( a − n p n p ( 1 − p ) ) 【n重伯努利试验】f_A=\frac{n_A}{n},P(A)=p\\ 令随机变量Y_n=\frac{n_A-np}{\sqrt{np(1-p)}},其分布函数F_n(x)\\ \lim_{n\rightarrow \infty}F_n(x)=\lim_{n\rightarrow \infty}P(\frac{n_A-np}{\sqrt{np(1-p)}}\leq x)=\int^x_{-\infty}\frac{1}{\sqrt{2\pi}}e^{-\frac{t^2}{2}}dt=\Phi(x)\\ \Longrightarrow X\thicksim N(np,np(1-p))(n\rightarrow \infty)\\ \Longrightarrow \left\{\begin{array}{c}P(X\leq x)\approx \Phi(\frac{x-np}{\sqrt{np(1-p)}})\\ P(a<X\leq b)\approx \Phi(\frac{b-np}{\sqrt{np(1-p)}})-\Phi(\frac{a-np}{\sqrt{np(1-p)}})\end{array}\right. 【n重伯努利试验】fA=nnA,P(A)=p令随机变量Yn=np(1−p)nA−np,其分布函数Fn(x)n→∞limFn(x)=n→∞limP(np(1−p)nA−np≤x)=∫−∞x2π1e−2t2dt=Φ(x)⟹X∼N(np,np(1−p))(n→∞)⟹⎩ ⎨ ⎧P(X≤x)≈Φ(np(1−p)x−np)P(a<X≤b)≈Φ(np(1−p)b−np)−Φ(np(1−p)a−np)