In the textbook, least square problem of simple regression was solved by partial derivative.
When introducing multiple regression, matrix approach was used to solve least square problems, which happens to produce same result with the partial derivative approach when solving simple regression. Other properties of least squares estimator can also be expressed in matrix form.
I was hoping to find the proof of the matrix approach, but I cannot find the proof in the textbook or online. Some keyword would be very helpful. Thank you!
