Consider a linear regression model:
where
for the residuals of the model. In the linear regression model with no constraints, we use the least squares method, hope and minimum residual sum of squares. which is
On this basis, plus a set of linear constraints:
written in matrix form:
Linear regression problem with constraint can be described as follows:
using Lagrange multiplier method:
for
derivative, after a series of calculations, resulting
And
are as follows:
wherein
is a parameter value without constraint.