7.1-exam3-2007¶
Quiz 3, 2007 · Section 7.1: Singular Values and Singular Vectors Source: exam3_1.pdf, p. 114
Problem¶
Problem 1 (25 points)
(a) Compute the singular value decomposition \(A = U\Sigma V^T\) for \(A = \begin{bmatrix} 0 & 1 \\ -1 & 1 \\ 1 & 0 \end{bmatrix}\).
(b) Find orthonormal bases for all four fundamental subspaces of \(A\).
Solution
(a) \(A^T A = \begin{bmatrix} 2 & -1 \\ -1 & 2 \end{bmatrix}\), with eigenvalues 3 and 1. Thus \(\sigma_1 = \sqrt{3}\) and \(\sigma_2 = 1\). \(v_1\) is a normal eigenvector corresponding to 3, so \(v_1 = \begin{bmatrix} 1/\sqrt{2} \\ -1/\sqrt{2} \end{bmatrix}\). \(v_2\) is a normal eigenvector corresponding to 1, so \(v_2 = \begin{bmatrix} 1/\sqrt{2} \\ 1/\sqrt{2} \end{bmatrix}\). \(u_1 = \frac{1}{\sqrt{3}}Av_1 = \begin{bmatrix} -1/\sqrt{6} \\ -2/\sqrt{6} \\ 1/\sqrt{6} \end{bmatrix}\); \(u_2 = Av_2 = \begin{bmatrix} 1/\sqrt{2} \\ 0 \\ 1/\sqrt{2} \end{bmatrix}\). For \(u_3\) we find a basis for the nullspace of \(A^T\): \(u_3 = \begin{bmatrix} -1/\sqrt{3} \\ 1/\sqrt{3} \\ 1/\sqrt{3} \end{bmatrix}\). Thus \[ A = \underbrace{\begin{bmatrix} -1/\sqrt{6} & 1/\sqrt{2} & -1/\sqrt{3} \\ -2/\sqrt{6} & 0 & 1/\sqrt{3} \\ 1/\sqrt{6} & 1/\sqrt{2} & 1/\sqrt{3} \end{bmatrix}}_{U} \underbrace{\begin{bmatrix} \sqrt{3} & 0 \\ 0 & 1 \\ 0 & 0 \end{bmatrix}}_{\Sigma} \underbrace{\begin{bmatrix} 1/\sqrt{2} & -1/\sqrt{2} \\ 1/\sqrt{2} & 1/\sqrt{2} \end{bmatrix}}_{V^T}. \] (b) Rowspace: \(v_1\) and \(v_2\), i.e. \(\begin{bmatrix} 1/\sqrt{2} \\ -1/\sqrt{2} \end{bmatrix}, \begin{bmatrix} 1/\sqrt{2} \\ 1/\sqrt{2} \end{bmatrix}\). Nullspace: 0. Columnspace: \(u_1\) and \(u_2\), i.e. \(\begin{bmatrix} -1/\sqrt{6} \\ -2/\sqrt{6} \\ 1/\sqrt{6} \end{bmatrix}, \begin{bmatrix} 1/\sqrt{2} \\ 0 \\ 1/\sqrt{2} \end{bmatrix}\). Left nullspace: \(u_3\), i.e. \(\begin{bmatrix} -1/\sqrt{3} \\ 1/\sqrt{3} \\ 1/\sqrt{3} \end{bmatrix}\).Keywords: singular value decomposition, SVD, fundamental subspaces, orthonormal basis