请教多元线性回归C++算法或程序

多元线性回归大含袜C++程序滚激：

Dim a() As String, b() As String

Private Sub Command1_Click()

Dim temp1 As String

Dim i As Integer, k As Integer

temp1 = Text1.Text

a = Split(temp1, ",")

lenolds = Len("->")

For i = 0 To UBound(a)

s = a(i)

j = InStr(s, "->")

Do While j >0

ReDim Preserve b(i, k)

b(i, k) = Val(Left(s, j + lenolds))

s = Left(s, j - 1) + "->" + Mid(s, j + lenolds + 1)

k = k + 1

j = InStr(s, "老侍->")

Loop

ReDim Preserve b(i, k)

b(i, k) = s

Next i

For i = 0 To UBound(a)

For j = 0 To k

Text2.Text = Text2.Text + b(i, j)

Next j

Next i

End Sub

题主的（缓码4自变量1因变量）多元函数的拟合matlab程序。可以用nlinfit非线性回归函数来做段察（也可以用lsqcurvefit函数）。实现代码：

x1=[0.55 0.65 0.65 0.65 0.65 0.55 0.55 0.55 0.65]'

x2=[1.6 1.6 1.4 1.6 1.4 1.4 1.6 1.4 1.4]'

x3=[20 20 20 10 10 20 20 20 20]'

x4=[10 10 10 10 10 10 12 12 12]'

Y=[1.848 3.145 3.337 3.022 3.188 1.97 1.63 1.621 2.534]'

X=[x1 x2 x3 x4]

n=length(x1)

a0=rand(1,5)

func=@(a,X)(a(1)+a(2)*X(:,1)+a(3)*X(:,2)+a(4)*X(:,3)+a(5)*X(:,4))

[a,r,J] = nlinfit(X,Y,func,a0)a

Y1=func(a,X)

[Y  Y1]

运行上述代码，可以得到

a1= -1.52778571412534；a2= 11.4823809522886；a3=  -0.359047619087202 ；a4= 0.00441190476147387；a5=-0.23363095238449

多元函数表达式，y=a1+a2*x1+a3*x2+a4*x3+a5*x4

nlinfit与lsqcurvefit两者区别并不扰燃哪太大，前者用回归的方法来求解，而后者用最小二乘法来求解，两者都可以用于非线性函数和线性函数。

---