高级优化理论与方法(精华版)
前言
这门课已经结课,马上就要考试了。在这里,我把一个学期的内容都整理一下,方便复习。这里就略过定义等基础内容,只有大致框架及重点内容。
无约束优化问题
必要条件和充分条件
FONC
FONC:一阶必要条件。若 x ∗ x^* x∗为最小值点,则 d T ∇ f ( x ∗ ) ≥ 0 d^T \nabla f(x^*)\geq0 dT∇f(x∗)≥0。若 x ∗ x^* x∗为内点,则 d T ∇ f ( x ∗ ) = 0 d^T\nabla f(x^*)=0 dT∇f(x∗)=0。
SONC
SONC:二阶必要条件。若 x ∗ x^* x∗为最小值点,且 d T ∇ f ( x ∗ ) = 0 d^T\nabla f(x^*)=0 dT∇f(x∗)=0,则 d T F ( x ∗ ) d ≥ 0 d^T F(x^*)d\geq0 dTF(x∗)d≥0。
SOSC
SOSC:二阶充分条件。若 x ∗ x^* x∗满足 { ∇ f ( x ∗ ) = 0 F ( x ∗ ) > 0 \begin{cases} \nabla f(x^*)=0\\ F(x^*)>0 \end{cases} {∇f(x∗)=0F(x∗)>0,则 x ∗ x^* x∗是局部最小值点。
应用
求无约束优化问题的步骤:
计算
∇
f
(
x
)
\nabla f(x)
∇f(x)和
F
(
x
)
F(x)
F(x),找到满足
{
∇
f
(
x
∗
)
=
0
F
(
x
∗
)
>
0
\begin{cases} \nabla f(x^*)=0\\ F(x^*)>0 \end{cases}
{∇f(x∗)=0F(x∗)>0的点
x
∗
x^*
x∗。
例子
f
(
x
)
=
x
1
2
+
x
2
2
f(x)=x_1^2+x_2^2
f(x)=x12+x22
∇
f
(
x
)
=
[
2
x
1
,
2
x
2
]
T
\nabla f(x)=[2x_1,2x_2]^T
∇f(x)=[2x1,2x2]T
H ( x ) = [ 2 0 0 2 ] > 0 H(x)=\begin{bmatrix} 2 & 0 \\ 0 & 2 \end{bmatrix}>0 H(x)=[2002]>0
x ∗ = [ 0 , 0 ] T x^*=[0,0]^T x∗=[0,0]T
算法
一维搜索算法
梯度下降法
公式
x k + 1 = x k − α k ∇ f ( x k ) x^{k+1}=x^k-\alpha^k \nabla f(x^k) xk+1=xk−αk∇f(xk), 其中 α k = a r g m i n f ( x k − α k ∇ f ( x k ) ) \alpha^k=argmin f(x^k-\alpha^k \nabla f(x^k)) αk=argminf(xk−αk∇f(xk))
若 f ( x ) f(x) f(x)为二次函数,即 f ( x ) = 1 2 x T Q x − b x f(x)=\frac{1}{2} x^TQx-bx f(x)=21xTQx−bx ,则 α k = g k T ⋅ g k g k T Q g k \alpha^k =\frac{{g^k}^T\cdot g^k}{{g^k}^TQg^k} αk=gkTQgkgkT⋅gk,其中 g k = ∇ f ( x k ) = Q x k − b g^k=\nabla f(x^k)=Qx^k-b gk=∇f(xk)=Qxk−b
应用
f
(
x
)
=
x
1
2
+
x
2
2
⇒
f
(
x
)
=
1
2
x
T
Q
x
−
b
x
f(x)=x_1^2+x_2^2 \Rightarrow f(x)=\frac{1}{2} x^TQx-bx
f(x)=x12+x22⇒f(x)=21xTQx−bx
Q
=
[
2
0
0
2
]
Q=\begin{bmatrix} 2&0 \\ 0&2 \end{bmatrix}
Q=[2002]
b = [ 0 0 ] b=\begin{bmatrix} 0 \\ 0 \end{bmatrix} b=[00]
∇ f ( x ) = [ 2 x 1 2 x 2 ] \nabla f(x)=\begin{bmatrix} 2x_1 \\ 2x_2 \end{bmatrix} ∇f(x)=[2x12x2]
F ( x ) = [ 2 0 0 2 ] F(x)=\begin{bmatrix} 2&0 \\ 0&2 \end{bmatrix} F(x)=[2002]
x 0 = [ 1 1 ] x^0=\begin{bmatrix} 1 \\ 1 \end{bmatrix} x0=[11]
x k + 1 = x k − α k g k x^{k+1}=x^k-\alpha^kg^k xk+1=xk−αkgk
x 1 = x 0 − α 0 g 0 x^1=x^0-\alpha^0g^0 x1=x0−α0g0
g 0 = ∇ f ( x 0 ) = [ 2 2 ] g^0=\nabla f(x^0)=\begin{bmatrix} 2 \\ 2 \end{bmatrix} g0=∇f(x0)=[22]
α 0 = g 0 T g 0 g 0 T Q g 0 = 8 [ 4 , 4 ] ⋅ [ 2 , 2 ] T = 8 16 = 1 2 \alpha^0=\frac{{g^0}^Tg^0}{{g^0}^TQg^0}=\frac{8}{[4,4]\cdot [2,2]^T}=\frac{8}{16}=\frac{1}{2} α0=g0TQg0g0Tg0=[4,4]⋅[2,2]T8=168=21
x 1 = [ 1 , 1 ] T − 1 2 [ 2 , 2 ] T = [ 0 , 0 ] T x^1=[1,1]^T-\frac{1}{2}[2,2]^T=[0,0]^T x1=[1,1]T−21[2,2]T=[0,0]T
g 1 = ∇ f ( x 1 ) = [ 0 0 ] g^1=\nabla f(x^1)=\begin{bmatrix} 0 \\ 0 \end{bmatrix} g1=∇f(x1)=[00]
牛顿法
一元函数: x k + 1 = x k − f ′ ( x k ) f ′ ′ ( x k ) x^{k+1}=x^k-\frac{f'(x^k)}{f''(x^k)} xk+1=xk−f′′(xk)f′(xk)
多元函数: x k + 1 = x k − [ F ( x k ) ] − 1 ∇ f ( x k ) x^{k+1}=x^k-[F(x^k)]^{-1}\nabla f(x^k) xk+1=xk−[F(xk)]−1∇f(xk)
针对二次函数的算法
共轭方向法
公式
二次函数情形下:
x
k
+
1
=
x
k
−
α
k
∇
f
(
x
k
)
x^{k+1}=x^k-\alpha^k \nabla f(x^k)
xk+1=xk−αk∇f(xk)
g k = ∇ f ( x k ) = Q x k − b g^k=\nabla f(x^k)=Qx^k-b gk=∇f(xk)=Qxk−b
α k = − g k T d k d k T Q d k \alpha^k=-\frac{{g^k}^Td_k}{d_k^TQd_k} αk=−dkTQdkgkTdk
x
k
+
1
=
x
k
+
α
k
d
k
x^{k+1}=x^k+\alpha^kd_k
xk+1=xk+αkdk
其中
d
0
,
d
1
,
⋯
,
d
n
d_0,d_1,\cdots,d_n
d0,d1,⋯,dn是一组共轭方向向量。
应用
求 f ( x ) = 1 2 x T [ 4 2 2 2 ] x − [ − 1 , 1 ] x f(x)=\frac{1}{2}x^T\begin{bmatrix} 4&2 \\ 2&2 \end{bmatrix}x-[-1,1]x f(x)=21xT[4222]x−[−1,1]x的最小值。
g ( x ) = [ 4 2 2 2 ] x − [ − 1 1 ] g(x)=\begin{bmatrix} 4&2 \\ 2&2 \end{bmatrix}x-\begin{bmatrix} -1 \\ 1 \end{bmatrix} g(x)=[4222]x−[−11]
选取 x 0 = [ 0 0 ] x^0=\begin{bmatrix} 0 \\ 0 \end{bmatrix} x0=[00]
选取 d 0 = [ 1 0 ] d_0=\begin{bmatrix} 1 \\ 0 \end{bmatrix} d0=[10]
d 0 T [ 4 2 2 2 ] d 1 = 0 ⇒ [ 4 2 ] d 1 = 0 d^T_0\begin{bmatrix} 4&2 \\ 2&2 \end{bmatrix}d_1=0\Rightarrow \begin{bmatrix} 4\\ 2 \end{bmatrix}d_1=0 d0T[4222]d1=0⇒[42]d1=0
选取 d 1 = [ − 3 8 3 4 ] d_1=\begin{bmatrix} -\frac{3}{8} \\ \frac{3}{4} \end{bmatrix} d1=[−8343]
g 0 = [ 1 − 1 ] g^0=\begin{bmatrix} 1 \\ -1 \end{bmatrix} g0=[1−1]
α 0 = − g 0 T d 0 d 0 T Q d 0 = − [ 1 , − 1 ] [ 1 0 ] [ 1 , 0 ] [ 4 2 2 2 ] [ 1 0 ] = − 1 4 \alpha^0=\frac{-{g^0}^Td_0}{d_0^TQd_0}=\frac{-[1,-1]\begin{bmatrix} 1 \\ 0 \end{bmatrix}}{[1,0]\begin{bmatrix} 4&2 \\ 2&2 \end{bmatrix}\begin{bmatrix} 1 \\ 0 \end{bmatrix}}=-\frac{1}{4} α0=d0TQd0−g0Td0=[1,0][4222][10]−[1,−1][10]=−41
x 1 = x 0 + α 0 d 0 = [ − 1 4 0 ] x^1=x^0+\alpha^0d_0=\begin{bmatrix} -\frac{1}{4} \\ 0 \end{bmatrix} x1=x0+α0d0=[−410]
g 1 = [ 0 − 3 2 ] g^1=\begin{bmatrix} 0 \\ -\frac{3}{2} \end{bmatrix} g1=[0−23]
α 1 = 2 \alpha^1=2 α1=2
x 2 = [ − 1 3 2 ] x^2=\begin{bmatrix} -1 \\ \frac{3}{2} \end{bmatrix} x2=[−123]
f ( x 2 ) = 0 f(x^2)=0 f(x2)=0
共轭梯度法
输入: f ( x ) = 1 2 x T Q x − b T x , x 0 f(x)=\frac{1}{2}x^TQx-b^Tx, x^0 f(x)=21xTQx−bTx,x0
- k = 0 k=0 k=0
- compute g 0 = ∇ f ( x 0 ) g^0=\nabla f(x^0) g0=∇f(x0), If g 0 = 0 g^0=0 g0=0, then stop; else d 0 = − g 0 d_0=-g^0 d0=−g0
- α k = − g k T d k d k T Q d k \alpha^k=-\frac{{g^k}^Td_k}{{d_k}^TQd_k} αk=−dkTQdkgkTdk
- x k + 1 = x k + α k d k x^{k+1}=x^k+\alpha^kd_k xk+1=xk+αkdk
- g k + 1 = ∇ f ( x k + 1 ) g^{k+1}=\nabla f(x^{k+1}) gk+1=∇f(xk+1), If g k + 1 = 0 g^{k+1}=0 gk+1=0 then stop;
- β k = g k + 1 T Q d k d k T Q d k \beta^k=\frac{{g^{k+1}}^TQd_k}{{d_k}^TQd_k} βk=dkTQdkgk+1TQdk
- d k + 1 = − g k + 1 + β k d k d^{k+1}=-g^{k+1}+\beta^kd_k dk+1=−gk+1+βkdk
- k k k++, goto 3
条件优化问题
线性规划
形式
min/max
f
(
x
)
=
c
T
x
=
∑
i
=
1
n
c
i
x
i
j
,
c
∈
R
n
,
x
∈
R
n
f(x)=c^Tx=\sum_{i=1}^nc_ix_{ij}, c\in \mathbb{R}^n, x \in \mathbb{R}^n
f(x)=cTx=∑i=1ncixij,c∈Rn,x∈Rn
s.t.
{
a
11
x
1
+
⋯
+
a
1
n
x
n
>
b
1
a
21
x
1
+
⋯
+
a
2
n
x
n
≤
b
2
⋯
a
m
1
x
1
+
⋯
+
a
m
n
x
n
≥
b
m
\begin{cases} a_{11}x_1+\cdots+a_{1n}x_n>b_1\\ a_{21}x_1+\cdots+a_{2n}x_n\leq b_2\\ \cdots\\ a_{m1}x_1+\cdots+a_{mn}x_n\geq b_m \end{cases}
⎩
⎨
⎧a11x1+⋯+a1nxn>b1a21x1+⋯+a2nxn≤b2⋯am1x1+⋯+amnxn≥bm
b
i
∈
R
,
∀
1
≤
i
≤
m
b_i\in\mathbb{R},\forall 1\leq i\leq m
bi∈R,∀1≤i≤m
a
i
j
∈
R
,
∀
1
≤
i
≤
n
,
1
≤
j
≤
m
a_{ij}\in\mathbb{R}, \forall 1\leq i\leq n, 1\leq j\leq m
aij∈R,∀1≤i≤n,1≤j≤m
单纯形法
算法
Simplex: Given an initial basis B
- Compute: [ A , b ] = [ B , D , b ] → [ I , Y , y 0 ] [A,b]=[B,D,b]\rightarrow [I,Y,y_0] [A,b]=[B,D,b]→[I,Y,y0]
- Compute: ∀ m + 1 ≤ j ≤ n : r j = c j − z j \forall m+1\leq j\leq n: r_j=c_j-z_j ∀m+1≤j≤n:rj=cj−zj
- If ∀ j : r j ≥ 0 \forall j: r_j\geq 0 ∀j:rj≥0, then return x = [ y 0 T , 0 ] T x=[y_0^T,0]^T x=[y0T,0]T
- else pick an q q q with r q < 0 r_q<0 rq<0
- If no
y
i
q
>
0
y_{iq}>0
yiq>0 for
1
≤
i
≤
m
1\leq i\leq m
1≤i≤m, then return “no solution”
else p = a r g m i n { y i 0 y i q : y i q > 0 } p=argmin\{\frac{y_{i0}}{y_{iq}}:y_{iq}>0\} p=argmin{yiqyi0:yiq>0} - Update [ I , Y , y 0 ] [I,Y,y_0] [I,Y,y0], goto 2
应用
max
7
x
1
+
6
x
2
7x_1+6x_2
7x1+6x2
s.t.
2
x
1
+
x
2
≤
3
2x_1+x_2\leq 3
2x1+x2≤3
x
1
+
4
x
2
≤
4
x_1+4x_2\leq 4
x1+4x2≤4
x
1
,
x
2
≥
0
x_1,x_2\geq 0
x1,x2≥0
⇒
\Rightarrow
⇒ min
−
7
x
1
−
6
x
2
-7x_1-6x_2
−7x1−6x2
s.t.
2
x
1
+
x
2
+
x
3
=
3
2x_1+x_2+x_3=3
2x1+x2+x3=3
x
1
+
4
x
2
+
x
4
=
4
x_1+4x_2+x_4=4
x1+4x2+x4=4
x
1
,
x
2
,
x
3
,
x
4
≥
0
x_1,x_2,x_3,x_4\geq 0
x1,x2,x3,x4≥0
[ A b c T 0 ] = [ 2 1 1 0 3 1 4 0 1 4 − 7 − 6 0 0 0 ] \begin{bmatrix} A&b\\ c^T&0 \end{bmatrix}=\begin{bmatrix} 2&1&1&0&3\\ 1&4&0&1&4\\ -7&-6&0&0&0 \end{bmatrix} [AcTb0]= 21−714−6100010340
B = [ a 3 , a 4 ] B=[a_3,a_4] B=[a3,a4]
x = [ 0 0 3 4 ] x=\begin{bmatrix} 0\\ 0\\ 3\\ 4 \end{bmatrix} x= 0034
r B T = [ − 7 , − 6 ] ⇒ q = 1 r_B^T=[-7,-6]\Rightarrow q=1 rBT=[−7,−6]⇒q=1
z 0 = 0 z_0=0 z0=0
p
=
a
r
g
m
i
n
{
y
i
0
y
i
q
:
y
i
q
>
0
}
p=argmin\{\frac{y_{i0}}{y_{iq}}:y_{iq}>0\}
p=argmin{yiqyi0:yiq>0}
3
2
\frac{3}{2}
23 v.s.
4
1
⇒
p
=
1
\frac{4}{1}\Rightarrow p=1
14⇒p=1
B ′ = [ a 1 , a 4 ] B'=[a_1,a_4] B′=[a1,a4]
[ 1 1 2 1 2 0 3 2 0 3 1 2 − 1 2 1 5 2 0 − 5 2 7 2 0 21 2 ] \begin{bmatrix} 1&\frac{1}{2}&\frac{1}{2}&0&\frac{3}{2}\\ 0&3\frac{1}{2}&-\frac{1}{2}&1&\frac{5}{2}\\ 0&-\frac{5}{2}&\frac{7}{2}&0&\frac{21}{2} \end{bmatrix} 10021321−2521−21270102325221
x = [ 3 2 , 0 , 0 , 5 2 ] T x=[\frac{3}{2},0,0,\frac{5}{2}]^T x=[23,0,0,25]T
z 0 = − 21 2 z_0=-\frac{21}{2} z0=−221
r D T = [ − 5 2 , 7 2 ] ⇒ q = 2 r_D^T=[-\frac{5}{2},\frac{7}{2}]\Rightarrow q=2 rDT=[−25,27]⇒q=2
y i 0 y i q : 3 2 1 2 = 3 \frac{y_{i0}}{y_{iq}}: \frac{\frac{3}{2}}{\frac{1}{2}}=3 yiqyi0:2123=3 v.s. 5 2 7 2 = 5 7 ⇒ p = 2 \frac{\frac{5}{2}}{\frac{7}{2}}=\frac{5}{7}\Rightarrow p=2 2725=75⇒p=2
[ 1 0 4 7 − 1 7 8 7 0 1 − 1 7 2 7 5 7 0 0 22 7 5 7 86 7 ] \begin{bmatrix} 1&0&\frac{4}{7}&-\frac{1}{7}&\frac{8}{7}\\ 0&1&-\frac{1}{7}&\frac{2}{7}&\frac{5}{7}\\ 0&0&\frac{22}{7}&\frac{5}{7}&\frac{86}{7} \end{bmatrix} 10001074−71722−7172757875786
B = [ a 1 , a 2 ] , x = [ 8 7 , 5 7 , 0 , 0 ] T , z = − 86 7 , r D T = [ 22 7 , 5 7 ] ≥ 0 B=[a_1,a_2],x=[\frac{8}{7},\frac{5}{7},0,0]^T,z=-\frac{86}{7},r_D^T=[\frac{22}{7},\frac{5}{7}]\geq 0 B=[a1,a2],x=[78,75,0,0]T,z=−786,rDT=[722,75]≥0
x x x 达到最优 !
非线性条件优化问题
约束条件只有等式的情形
min
f
(
x
)
=
0
f(x)=0
f(x)=0
s.t.
h
(
x
)
=
0
h(x)=0
h(x)=0
拉格朗日条件
公式
拉格朗日函数: l ( x , λ ) = f ( x ) + λ T h ( x ) l(x,\lambda)=f(x)+\lambda^Th(x) l(x,λ)=f(x)+λTh(x)
D l ( x ∗ , λ ∗ ) = 0 ⇒ { D x l ( x ∗ , λ ∗ ) = 0 D λ l ( x ∗ , λ ∗ ) = 0 Dl(x^*,\lambda^*)=0\Rightarrow \begin{cases} D_xl(x^*,\lambda^*)=0\\ D_{\lambda}l(x^*,\lambda^*)=0 \end{cases} Dl(x∗,λ∗)=0⇒{Dxl(x∗,λ∗)=0Dλl(x∗,λ∗)=0
应用
已知长方体的表面积为
A
A
A,求体积的最大值。
max
x
1
x
2
x
3
x_1x_2x_3
x1x2x3
s.t.
x
1
x
2
+
x
2
x
3
+
x
1
x
3
=
A
2
(
A
>
0
)
x_1x_2+x_2x_3+x_1x_3=\frac{A}{2}(A>0)
x1x2+x2x3+x1x3=2A(A>0)
f
(
x
)
=
−
x
1
x
2
x
3
,
h
(
x
)
=
x
1
x
2
+
x
2
x
3
+
x
1
x
3
−
A
2
f(x)=-x_1x_2x_3,h(x)=x_1x_2+x_2x_3+x_1x_3-\frac{A}{2}
f(x)=−x1x2x3,h(x)=x1x2+x2x3+x1x3−2A
∇
f
(
x
)
=
[
−
x
2
x
3
,
−
x
1
x
3
,
−
x
1
x
2
]
T
\nabla f(x)=[-x_2x_3,-x_1x_3,-x_1x_2]^T
∇f(x)=[−x2x3,−x1x3,−x1x2]T
∇
h
(
x
)
=
[
x
2
+
x
3
,
x
1
+
x
3
,
x
1
+
x
2
]
T
\nabla h(x)=[x_2+x_3,x_1+x_3,x_1+x_2]^T
∇h(x)=[x2+x3,x1+x3,x1+x2]T
All feasible solutions are regular.
λ
∈
R
\lambda\in\mathbb{R}
λ∈R
{
∇
f
(
x
)
+
λ
∇
h
(
x
)
=
0
h
(
x
)
=
0
⇒
{
x
2
x
3
−
λ
(
x
2
+
x
3
)
=
0
x
1
x
3
−
λ
(
x
1
+
x
3
)
=
0
x
1
x
2
−
λ
(
x
1
+
x
2
)
=
0
x
1
x
2
+
x
2
x
3
+
x
1
x
3
−
A
2
=
0
\begin{cases} \nabla f(x)+\lambda \nabla h(x)=0\\ h(x)=0 \end{cases}\Rightarrow \begin{cases} x_2x_3-\lambda(x_2+x_3)=0\\ x_1x_3-\lambda(x_1+x_3)=0\\ x_1x_2-\lambda(x_1+x_2)=0\\ x_1x_2+x_2x_3+x_1x_3-\frac{A}{2}=0 \end{cases}
{∇f(x)+λ∇h(x)=0h(x)=0⇒⎩
⎨
⎧x2x3−λ(x2+x3)=0x1x3−λ(x1+x3)=0x1x2−λ(x1+x2)=0x1x2+x2x3+x1x3−2A=0
当 x 1 = x 2 = x 3 = A 6 x_1=x_2=x_3=\sqrt{\frac{A}{6}} x1=x2=x3=6A时,取到最值
约束条件包含不等式的情形
min
f
(
x
)
f(x)
f(x)
s.t.
h
(
x
)
=
0
h(x)=0
h(x)=0
g
(
x
)
≤
0
g(x)\leq 0
g(x)≤0
KKT条件
公式
- μ ∗ ≥ 0 \mu^*\geq 0 μ∗≥0
- D f ( x ∗ ) + λ ∗ T D h ( x ∗ ) + μ ∗ T D g ( x ∗ ) = 0 Df(x^*)+{\lambda^*}^TDh(x^*)+{\mu^*}^TDg(x^*)=0 Df(x∗)+λ∗TDh(x∗)+μ∗TDg(x∗)=0
- μ ∗ T g ( x ∗ ) = 0 {\mu^*}^Tg(x^*)=0 μ∗Tg(x∗)=0
应用
min
−
400
R
(
10
+
R
)
2
-\frac{400R}{(10+R)^2}
−(10+R)2400R
s.t.
−
R
≤
0
-R\leq 0
−R≤0
∇ f ( R ) = − 400 ( 10 − R ) ( 10 + R ) 3 \nabla f(R)=-\frac{400(10-R)}{(10+R)^3} ∇f(R)=−(10+R)3400(10−R)
{ μ ≥ 0 D f ( x ∗ ) + λ ∗ T D h ( x ∗ ) + μ ∗ T D g ( x ∗ ) = 0 μ T g ( x ) = 0 g ( x ) ≤ 0 h ( x ) = 0 \begin{cases} \mu\geq 0\\ Df(x^*)+{\lambda^*}^TDh(x^*)+{\mu^*}^TDg(x^*)=0\\ \mu^T g(x)=0\\ g(x)\leq 0\\ h(x)=0 \end{cases} ⎩ ⎨ ⎧μ≥0Df(x∗)+λ∗TDh(x∗)+μ∗TDg(x∗)=0μTg(x)=0g(x)≤0h(x)=0
⇒ { μ ≥ 0 − 400 ( 10 − R ) ( 10 + R ) 3 − μ = 0 μ R = 0 R ≥ 0 \Rightarrow \begin{cases} \mu\geq 0\\ -\frac{400(10-R)}{(10+R)^3}-\mu=0\\ \mu R=0\\ R\geq 0 \end{cases} ⇒⎩ ⎨ ⎧μ≥0−(10+R)3400(10−R)−μ=0μR=0R≥0
If
μ
>
0
\mu>0
μ>0, then
R
=
0
,
μ
=
−
4
R=0,\mu=-4
R=0,μ=−4(✕)
If
μ
=
0
⇒
R
=
10
\mu=0\Rightarrow R=10
μ=0⇒R=10(✓ )
考试押题
一共五道大题,我也押五道题。
FONC,SONC,SOSC的应用。
梯度下降法。
共轭方向法。
单纯形法。
KKT条件。
总结
我把这个学期讲的重点全部整理在这里了。其实这门课程的内容非常多,其中包括了很多难度比较大的内容,以及为了知识体系的完整性补充的内容,那些内容这里全都没有放上来。这里只挑选了一些难度不大,讲解较为详细,应用较广的内容。也祝所有选这门课的同学考高分!