文章目录
引言
这数一全部内容太多了,放在一篇文章里的话,要编辑就很困难,就把线代和概率放在这篇文章里吧。
二、线代
施密特正交化
把一组线性无关的向量组转化为一组两两正交且规范的向量组的过程,称为施密特正交化。
设 α 1 , α 2 , ⋯ , α n \pmb{\alpha_1,\alpha_2,\cdots,\alpha_n} α1,α2,⋯,αn 线性无关,其正交化过程为:
(1)正交化 l e t β 1 = α 1 , β 2 = α 2 − ( α 2 , β 1 ) ( β 1 , β 1 ) β 1 β n = α n − ( α n , β 1 ) ( β 1 , β 1 ) β 1 − ( α n , β 2 ) ( β 2 , β 2 ) β 2 − ⋯ − ( α n , β n − 1 ) ( β n − 1 , β n − 1 ) β n − 1 let\space \pmb{\beta_1=\alpha_1,\beta_2=\alpha_2-\frac{(\alpha_2,\beta_1)}{(\beta_1,\beta_1)}\beta_1}\\ \pmb{\beta_n=\alpha_n-\frac{(\alpha_n,\beta_1)}{(\beta_1,\beta_1)}\beta_1-\frac{(\alpha_n,\beta_2)}{(\beta_2,\beta_2)}\beta_2}-\cdots-\pmb{\frac{(\alpha_n,\beta_{n-1})}{(\beta_{n-1},\beta_{n-1})}\beta_{n-1}} let β1=α1,β2=α2−(β1,β1)(α2,β1)β1βn=αn−(β1,β1)(αn,β1)β1−(β2,β2)(αn,β2)β2−⋯−(βn−1,βn−1)(αn,βn−1)βn−1 则向量组 β 1 , β 2 , ⋯ , β n \pmb{\beta_1,\beta_2,\cdots,\beta_n} β1,β2,⋯,βn 两两正交。
(2)规范化。各自除以各自的模即可。
分块矩阵
首先是行列式,有以下三个结论:
(1) ∣ A 1 A 2 ⋱ A n ∣ = ∣ A 1 ∣ ⋅ ∣ A 2 ∣ ⋯ ∣ A n ∣ . \begin{vmatrix} \pmb{A_1} & & & \\ & \pmb{A_2} & & \\ & & \ddots & \\ & & & \pmb{A_n}\end{vmatrix}=|\pmb{A_1}|\cdot|\pmb{A_2}|\cdots|\pmb{A_n}|. A1A2⋱An =∣A1∣⋅∣A2∣⋯∣An∣.
(2) ∣ A C O B ∣ = ∣ A O O B ∣ = ∣ A ∣ ⋅ ∣ B ∣ . \begin{vmatrix} \pmb{A} & \pmb{C}\\ \pmb{O}& \pmb{B} \end{vmatrix}=\begin{vmatrix} \pmb{A} & \pmb{O}\\ \pmb{O}& \pmb{B} \end{vmatrix}=|\pmb{A}|\cdot|\pmb{B}|. AOCB = AOOB =∣A∣⋅∣B∣.
(3)设 A , B \pmb{A,B} A,B 分别为 m , n m,n m,n 阶方阵,则有 ∣ O A B O ∣ = ( − 1 ) m n ∣ A ∣ ⋅ ∣ B ∣ . \begin{vmatrix} \pmb{O} & \pmb{A}\\ \pmb{B}& \pmb{O} \end{vmatrix}=(-1)^{mn}|\pmb{A}|\cdot|\pmb{B}|. OBAO =(−1)mn∣A∣⋅∣B∣.
然后是转置的结论: [ A B C D ] T = [ A T C T B T D T ] . \begin{bmatrix} \pmb{A} & \pmb{B}\\ \pmb{C}& \pmb{D} \end{bmatrix}^T=\begin{bmatrix} \pmb{A^T} & \pmb{C^T}\\ \pmb{B^T}& \pmb{D^T} \end{bmatrix}. [ACBD]T=[ATBTCTDT].
接着是逆矩阵的结论: [ A O O B ] − 1 = [ A − 1 O O B − 1 ] , [ O A B O ] − 1 = [ O B − 1 A − 1 O ] . \begin{bmatrix} \pmb{A} & \pmb{O}\\ \pmb{O}& \pmb{B} \end{bmatrix}^{-1}=\begin{bmatrix} \pmb{A^{-1}} & \pmb{O}\\ \pmb{O}& \pmb{B^{-1}} \end{bmatrix},\begin{bmatrix} \pmb{O} & \pmb{A}\\ \pmb{B}& \pmb{O} \end{bmatrix}^{-1}=\begin{bmatrix} \pmb{O} & \pmb{B^{-1}}\\ \pmb{A^{-1}}& \pmb{O} \end{bmatrix}. [AOOB]−1=[A−1OOB−1],[OBAO]−1=[OA−1B−1O].
转置、逆、伴随之间的运算
对可逆矩阵,转置、逆和伴随可以随意交换顺序,即 ( A − 1 ) T = ( A T ) − 1 , ( A ∗ ) − 1 = ( A − 1 ) ∗ , ( A ∗ ) T = ( A T ) ∗ . (\pmb{A}^{-1})^T=(\pmb{A}^{T})^{-1},(\pmb{A}^{*})^{-1}=(\pmb{A}^{-1})^{*},(\pmb{A}^{*})^T=(\pmb{A}^{T})^*. (A−1)T=(AT)−1,(A∗)−1=(A−1)∗,(A∗)T=(AT)∗.
关于秩
定义
矩阵的秩的定义:
设 A \pmb{A} A 是 m × n m\times n m×n 矩阵,从中任取 r r r 行 r r r 列,元素按照原有次序构成的 r r r 阶行列式,称为矩阵 A \pmb{A} A 的 r r r 阶子式。若 矩阵 A \pmb{A} A 中至少有一个 r r r 阶子式不为零,但所有 r + 1 r+1 r+1 阶子式(可能没有)均为零,称 r r r 为矩阵 A \pmb{A} A 的秩。
向量组秩的定义:
设 α 1 , α 2 , ⋯ , α n \pmb{\alpha_1,\alpha_2,\cdots,\alpha_n} α1,α2,⋯,αn 为一组向量,若其存在 r r r 个向量线性无关,且任意 r + 1 r+1 r+1 个向量(不一定有)一定线性相关,称这 r r r 个线性无关的向量构成的向量组为 α 1 , α 2 , ⋯ , α n \pmb{\alpha_1,\alpha_2,\cdots,\alpha_n} α1,α2,⋯,αn 的极大线性无关组,极大线性无关组所含向量的个数,称为向量组的秩。
性质
矩阵的秩有如下性质: r ( A ) = r ( A T ) = r ( A A T ) = r ( A T A ) . [ r ( A ) + r ( B ) − n ] ≤ r ( A + B ) ≤ r ( A ) + r ( B ) . r ( A B ) ≤ min { r ( A ) , r ( B ) } . i f A B = O , t h e n , r ( A ) + r ( B ) ≤ n . i f ∣ P ∣ , ∣ Q ∣ ≠ 0 , r ( A ) = r ( P A ) = r ( A Q ) = r ( P A Q ) . r ( A ∗ ) = { n r ( A ) = n 1 r ( A ) = n − 1 0 r ( A ) < n − 1 , ( n ≥ 2 ) . l e t A m × n , B m × s , t h e n , max { r ( A ) , r ( A ) } ≤ r ( A ⋮ B ) ≤ r ( A ) + r ( B ) . α , β ≠ 0 , r ( A ) = 1 ⟺ A = α β T . r ( A O O B ) = r ( A ) + r ( A ) . r(\pmb{A})=r(\pmb{A}^T)=r(\pmb{A}\pmb{A}^T)=r(\pmb{A}^T\pmb{A}).\\ [r(\pmb{A})+r(\pmb{B})-n]\leq r(\pmb{A}+\pmb{B})\leq r(\pmb{A})+r(\pmb{B}). \\ r(\pmb{AB})\leq \min\{r(\pmb{A}),r(\pmb{B})\}. \\ if\space \pmb{AB=O},then\space ,r(\pmb{A})+r(\pmb{B})\leq n. \\ if\space |\pmb{P}|,|\pmb{Q}|\ne0,r(\pmb{A})=r(\pmb{PA})=r(\pmb{AQ})=r(\pmb{PAQ}).\\ r(\pmb{A}^*)=\begin{cases} n&r(\pmb{A})=n\\ 1&r(\pmb{A})=n-1\\ 0&r(\pmb{A})<n-1 \end{cases},(n\geq2).\\ let\space \pmb{A}{m\times n},\pmb{B}{m\times s},then,\max\{r(\pmb{A}),r(\pmb{A})\}\leq r(\pmb{A}\space\vdots \space B)\leq r(\pmb{A})+r(\pmb{B}). \\ \pmb{\alpha,\beta\ne 0},r(\pmb{A})=1 \pmb{\Longleftrightarrow} \pmb{A}=\pmb{\alpha\beta}^T.\\ r\begin{pmatrix} \pmb{A} & \pmb{O} \\ \pmb{O}& \pmb{B}\end{pmatrix}=r(\pmb{A})+r(\pmb{A}). r(A)=r(AT)=r(AAT)=r(ATA).[r(A)+r(B)−n]≤r(A+B)≤r(A)+r(B).r(AB)≤min{r(A),r(B)}.if AB=O,then ,r(A)+r(B)≤n.if ∣P∣,∣Q∣=0,r(A)=r(PA)=r(AQ)=r(PAQ).r(A∗)=⎩ ⎨ ⎧n10r(A)=nr(A)=n−1r(A)<n−1,(n≥2).let Am×n,Bm×s,then,max{r(A),r(A)}≤r(A ⋮ B)≤r(A)+r(B).α,β=0,r(A)=1⟺A=αβT.r(AOOB)=r(A)+r(A).
三、概统
常见分布的期望及方差
{ 分布 ‾ 分布律或概率密度 ‾ 数学期望 ‾ 方差 ‾ ( 0 − 1 )分布 P { X = k } = p k ( 1 − p ) 1 − k , k = 0 , 1 p p ( 1 − p ) 二项分布 P { X = k } = C n k p k ( 1 − p ) n − k , k = 0 ⋯ n n p n p ( 1 − p ) 泊松分布 P { X = k } = λ k k ! e − λ , k = 0 , 1 , 2 , ⋯ λ λ 正态分布 f ( x ) = 1 2 π σ E X P ( − ( x − μ ) 2 2 σ 2 ) μ σ 2 几何分布 P { X = k } = ( 1 − p ) k − 1 p , k = 1 , 2 , ⋯ 1 / p ( 1 − p ) / p 2 \begin{cases}\underline{分布}&\underline{分布律或概率密度}&\underline{数学期望}&\underline{方差}\\ (0-1)分布&P\{X=k\}=p^k(1-p)^{1-k},k=0,1&p&p(1-p)\\ 二项分布& P\{X=k\}=C_n^kp^k(1-p)^{n-k},k=0\cdots n&np&np(1-p)\\ 泊松分布&P\{X=k\}=\frac{\lambda^k}{k!}e^{-\lambda},k=0,1,2,\cdots&\lambda&\lambda \\ 正态分布 & f(x)=\frac{1}{\sqrt{2\pi}\sigma}E XP(-\frac{(x-\mu)^2}{2\sigma^2})&\mu&\sigma^2\\ 几何分布&P\{X=k\}=(1-p)^{k-1}p,k=1,2,\cdots&1/p&(1-p)/p^2\end{cases} ⎩ ⎨ ⎧分布(0−1)分布二项分布泊松分布正态分布几何分布分布律或概率密度P{X=k}=pk(1−p)1−k,k=0,1P{X=k}=Cnkpk(1−p)n−k,k=0⋯nP{X=k}=k!λke−λ,k=0,1,2,⋯f(x)=2π σ1EXP(−2σ2(x−μ)2)P{X=k}=(1−p)k−1p,k=1,2,⋯数学期望pnpλμ1/p方差p(1−p)np(1−p)λσ2(1−p)/p2 均匀分布: f ( x ) = { 1 / ( b − a ) , a < x < b 0 , e l s e , E ( X ) = a + b 2 , D ( X ) = ( b − a ) 2 12 . f(x)=\begin{cases} 1/(b-a),&a<x<b \\ 0,&else \end{cases},E(X)=\frac{a+b}{2},D(X)=\frac{(b-a)^2}{12}. f(x)={1/(b−a),0,a<x<belse,E(X)=2a+b,D(X)=12(b−a)2. 指数分布: f ( x ) = { λ e − λ x , x > 0 0 , e l s e , E ( X ) = 1 λ , D ( X ) = 1 λ 2 . f(x)=\begin{cases} \lambda e^{-\lambda x},&x>0 \\ 0,&else \end{cases},E(X)=\frac{1}{\lambda},D(X)=\frac{1}{\lambda^2}. f(x)={λe−λx,0,x>0else,E(X)=λ1,D(X)=λ21.