\(\newcommand{\bx}{\textbf{x}} \newcommand{\bo}{\textbf{0}} \newcommand{\bv}{\textbf{v}} \newcommand{\bu}{\textbf{u}} \newcommand{\bq}{\textbf{q}} \newcommand{\by}{\textbf{y}} \newcommand{\bb}{\textbf{b}} \newcommand{\ba}{\textbf{a}} \newcommand{\grad}{\boldsymbol{\nabla}} \newcommand{\pd}[2]{\frac{\partial #1}{\partial #2}} \newcommand{\pdd}[2]{\frac{\partial^2 #1}{\partial #2^2}} \newcommand{\pddm}[3]{\frac{\partial^2 #1}{\partial #2 \partial #3}} \newcommand{\deriv}[2]{\frac{d #1}{d #2}} \newcommand{\lt}{ < } \newcommand{\gt}{ > } \newcommand{\amp}{ & } \)

Section5.3The QR Factorization

Our goal in this section is to improve the efficiency of solving the least squares problems via the normal equations. Recall that we want to find \(\bx\) such that \(A \bx = \bb\) but where \(\bb\) is not in the column space of \(A\). This necessitated the use of the normal equations \(A^T A \bx = A^T \bb\) to project \(\bb\) onto the column space of \(A\). This would be FAR more efficient if the columns of \(A\) were orthogonal (perpendicular) and normalized (unit vectors). Hence, our goal will be to take \(A\) and factor it into \(A = QR\) where the columns of \(Q\) are orthonormal (orthogonal and normalized) and \(R\) is an upper triangular matrix.