6.2-exam3-1999-b¶
Quiz 3, 1999 · Section 6.2: Diagonalizing a Matrix Source: exam3_1.pdf, pp. 31, 39
Problem¶
2 (25 pts.) Suppose a 3 by 3 real symmetric matrix \(A\) has eigenvalues 4, 4, 0.
(a) Find the determinant from the eigenvalues of \((2A - I)^{-1}\).
(b) True or false or not enough information: This matrix \(A\) has 3 independent eigenvectors and can be diagonalized.
(c) True or false or not enough information: The function \(x^TAx\) is never negative for any vector \(x\).
(d) True or false or not enough information: The matrix \(\frac{1}{4}A\) is a Markov matrix.
(e) If \(A\) has orthonormal eigenvectors \(q_1, q_2, q_3\) with \(\lambda = 4, 4, 0\), find a formula for \(A\) in terms of \(q_1, q_2, q_3\) using diagonalization.
Solution
(a) \((2A - I)^{-1}\) has eigenvalues \(\frac{1}{7}, \frac{1}{7}, -1\) (since \(2A - I\) has eigenvalues \(7, 7, -1\)), so the determinant is \[ -\frac{1}{49}. \] (b) True (\(A\) is given as symmetric, so it has 3 independent eigenvectors and can be diagonalized). (c) True (no negative eigenvalues; \(A\) is positive semidefinite, so \(x^TAx\) is never negative). (d) Not enough information: \(\frac{1}{4}A = \begin{bmatrix} 1 & 0 & 0 \\ 0 & 1 & 0 \\ 0 & 0 & 0 \end{bmatrix}\) is not Markov (the third column does not sum to 1), while \(\frac{1}{4}A = \begin{bmatrix} 1 & 0 & 0 \\ 0 & .5 & .5 \\ 0 & .5 & .5 \end{bmatrix}\) is Markov. (e) \[ A = Q\Lambda Q^T = \begin{bmatrix} q_1 & q_2 & q_3 \end{bmatrix} \begin{bmatrix} 4 & 0 & 0 \\ 0 & 4 & 0 \\ 0 & 0 & 0 \end{bmatrix} \begin{bmatrix} q_1^T \\ q_2^T \\ q_3^T \end{bmatrix} = 4q_1q_1^T + 4q_2q_2^T. \]Keywords: eigenvalues, symmetric matrix, diagonalization, positive semidefinite