Matlab 带约束的最小二乘法
我有37个线性方程组，其中36个变量采用矩阵形式：A x=b。（A有37行36列。）方程没有精确解，所以我使用了Matlab，使用Matlab 带约束的最小二乘法,matlab,math,curve-fitting,linear-programming,least-squares,Matlab,Math,Curve Fitting,Linear Programming,Least Squares,我有37个线性方程组，其中36个变量采用矩阵形式：A x=b。（A有37行36列。）方程没有精确解，所以我使用了Matlab，使用x=A\b 问题是，我也有一个条件，所有的元素X都应该是正的：对于所有的I，席> 0。code>x=A\b为某些元素提供负值。如何应用此约束 以下是我正在使用的A和b的具体值： A = [0.83 0.17 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
x=A\b以下是我正在使用的A和b的具体值：
A = [0.83 0.17 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0.02 0.63 0.17 0 0 0 0.18 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0.02 -0.37 0.17 0 0 0 0.18 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0.02 -0.37 0.17 0 0 0 0.18 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0.02 -0.37 0.17 0 0 0 0.18 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0.02 -0.2 0 0 0 0 0.18 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0.15 0 0 0 0 0 -0.32 0.17 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0.15 0 0 0 0 0.02 -0.52 0.17 0 0 0 0.18 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0.15 0 0 0 0 0.02 -0.52 0.17 0 0 0 0.18 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0.15 0 0 0 0 0.02 -0.52 0.17 0 0 0 0.18 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0.15 0 0 0 0 0.02 -0.52 0.17 0 0 0 0.18 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0.15 0 0 0 0 0.02 -0.33 0 0 0 0 0.18 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0.15 0 0 0 0 0 -0.32 0.17 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0.15 0 0 0 0 0.02 -0.52 0.17 0 0 0 0.18 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0.15 0 0 0 0 0.02 -0.52 0.17 0 0 0 0.18 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0.15 0 0 0 0 0.02 -0.52 0.17 0 0 0 0.18 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0.15 0 0 0 0 0.02 -0.52 0.17 0 0 0 0.18 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0.15 0 0 0 0 0.02 -0.35 0 0 0 0 0.18 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0.15 0 0 0 0 0 -0.32 0.17 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0.15 0 0 0 0 0.02 -0.52 0.17 0 0 0 0.18 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.15 0 0 0 0 0.02 -0.52 0.17 0 0 0 0.18 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.15 0 0 0 0 0.02 -0.52 0.17 0 0 0 0.18 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.15 0 0 0 0 0.02 -0.52 0.17 0 0 0 0.18 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.15 0 0 0 0 0.02 -0.35 0 0 0 0 0.18 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.15 0 0 0 0 0 -0.32 0.17 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.15 0 0 0 0 0.02 -0.52 0.17 0 0 0 0.18 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.15 0 0 0 0 0.02 -0.52 0.17 0 0 0 0.18 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.15 0 0 0 0 0.02 -0.52 0.17 0 0 0 0.18 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.15 0 0 0 0 0.02 -0.52 0.17 0 0 0 0.18 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.15 0 0 0 0 0.02 -0.35 0 0 0 0 0.018 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.15 0 0 0 0 0 -0.32 0.17 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.15 0 0 0 0 0.02 -0.34 0.17 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.15 0 0 0 0 0.02 -0.34 0.17 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.15 0 0 0 0 0.02 -0.34 0.17 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.15 0 0 0 0 0.02 -0.34 0.17
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.15 0 0 0 0 0.02 -0.17
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1];
b = [0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1]';
当x>0时，这将是一个线性优化问题。解决这个问题的最佳算法是单纯形算法。其思想是，每个线性方程aix=bi提供一条直线，这些直线的组合提供一个多边形/多面体，答案是该多面体/多边形的一个顶点。单纯形算法是非常标准的，有许多可用的函数和库可以计算它。当你有x>0的条件时，这将是一个线性优化问题。解决这个问题的最佳算法是单纯形算法。其思想是，每个线性方程aix=bi提供一条直线，这些直线的组合提供一个多边形/多面体，答案是该多面体/多边形的一个顶点。单纯形算法是非常标准的，有许多可用的函数和库可以进行计算。您希望在最小二乘意义下找到x到A x=b的近似解，即，您希望最小化 ||A x-b | | 2=x处的xT +btb-2xt在b 忽略常数项bTb，除以系数2，这符合a的形式，即最小化 1/2英尺高x+英尺高x 如果我们在A处选择H，在b处选择f=-
quadprogH = A' * A;
f = - A' * b;
x = quadprog(H, f)
quadprogAbx = quadprog(H, f, -eye(n), zeros(n, 1) + 0.001)
用A和b的实际数据补充问题后更新：使用代码
H = A' * A;
f = - A' * b;
n = size(A, 2);
x = quadprog(H, f, -eye(n), zeros(n, 1))
Minimum found that satisfies the constraints.
Optimization completed because the objective function is non-decreasing in
feasible directions, to within the default value of the function tolerance,
and constraints are satisfied to within the default value of the constraint tolerance.
<stopping criteria details>
x =
0.000380906335150292
3.90638261088393e-05
0.0111196970167585
0.0227055107206744
0.0318402514628274
0.0371743514880516
0.000800900221354844
0.00746652476710186
0.0180511534370576
0.0282423767946842
0.0362606972021829
0.0417582260990786
0.00860220929402253
0.0174105435824309
0.0265771677458008
0.0343071472371469
0.0395176470725881
0.0419494410289298
0.0187719294637544
0.0268976053211278
0.0336818044612046
0.0382365751296441
0.0398823076542831
0.0391016682549663
0.0279383031707377
0.0339393563379992
0.0377917413001034
0.0382731422972829
0.0338557405807941
0.0217568643500703
0.0343698083354502
0.0381554349806972
0.0392353941260779
0.0368010570888738
0.031271868401718
0.0258232230013864
找到满足约束条件的最小值。
优化已完成，因为目标函数在
可行方向，在功能公差的默认值范围内，
和约束满足在约束公差的默认值范围内。
x=
0.000380906335150292
3.90638261088393e-05
0.0111196970167585
0.0227055107206744
0.0318402514628274
0.0371743514880516
0.000800900221354844
0.00746652476710186
0.0180511534370576
0.0282423767946842
0.0362606972021829
0.0417582260990786
0.00860220929402253
0.0174105435824309
0.0265771677458008
0.0343071472371469
0.0395176470725881
0.0419494410289298
0.0187719294637544
0.0268976053211278
0.0336818044612046
0.0382365751296441
0.0398823076542831
0.0391016682549663
0.0279383031707377
0.0339393563379992
0.0377917413001034
0.0382731422972829
0.0338557405807941
0.0217568643500703
0.0343698083354502
0.0381554349806972
0.0392353941260779
0.0368010570888738
0.031271868401718
0.0258232230013864
quadprogH = A' * A;
f = - A' * b;
x = quadprog(H, f)
quadprogAbx = quadprog(H, f, -eye(n), zeros(n, 1) + 0.001)
用A和b的实际数据补充问题后更新：使用代码
H = A' * A;
f = - A' * b;
n = size(A, 2);
x = quadprog(H, f, -eye(n), zeros(n, 1))
Minimum found that satisfies the constraints.
Optimization completed because the objective function is non-decreasing in
feasible directions, to within the default value of the function tolerance,
and constraints are satisfied to within the default value of the constraint tolerance.
<stopping criteria details>
x =
0.000380906335150292
3.90638261088393e-05
0.0111196970167585
0.0227055107206744
0.0318402514628274
0.0371743514880516
0.000800900221354844
0.00746652476710186
0.0180511534370576
0.0282423767946842
0.0362606972021829
0.0417582260990786
0.00860220929402253
0.0174105435824309
0.0265771677458008
0.0343071472371469
0.0395176470725881
0.0419494410289298
0.0187719294637544
0.0268976053211278
0.0336818044612046
0.0382365751296441
0.0398823076542831
0.0391016682549663
0.0279383031707377
0.0339393563379992
0.0377917413001034
0.0382731422972829
0.0338557405807941
0.0217568643500703
0.0343698083354502
0.0381554349806972
0.0392353941260779
0.0368010570888738
0.031271868401718
0.0258232230013864
找到满足约束条件的最小值。
优化已完成，因为目标函数在
可行方向，在功能公差的默认值范围内，
和约束满足在约束公差的默认值范围内。
x=
0.000380906335150292
3.90638261088393e-05
0.0111196970167585
0.0227055107206744
0.0318402514628274
0.0371743514880516
0.000800900221354844
0.00746652476710186
0.0180511534370576
0.0282423767946842
0.0362606972021829
0.0417582260990786
0.00860220929402253
0.0174105435824309
0.0265771677458008
0