6.2-exam3-2005¶
Quiz 3, 2005 · Section 6.2: Diagonalizing a Matrix Source: exam3_1.pdf, pp. 79, 86-87
Problem¶
1 (30 pts.) (a) Find the eigenvalues and eigenvectors of the Markov matrix
\[
A = \begin{bmatrix} .9 & .4 \\ .1 & .6 \end{bmatrix}.
\]
(b) What is the limiting value of \(A^k \begin{bmatrix} 3 \\ 2 \end{bmatrix}\) as the power \(k\) goes to infinity?
(c) What does it mean to say that \(A\) is similar to \(B\)?
Is that 2 by 2 matrix \(A\) similar (yes or no) to its transpose \(B\)?
\[
B = \begin{bmatrix} .9 & .1 \\ .4 & .6 \end{bmatrix}
\]
Give a reason for your answer.
Solution
(a) Any Markov matrix has eigenvalue \(\lambda_1 = 1\); since the trace of \(A\) is 1.5, and the eigenvalues of a matrix add up to its trace, the second eigenvalue is \(\lambda_2 = .5\). To find the corresponding eigenvectors \(v_1\) and \(v_2\), we look at \(A - \lambda_1 I\) and \(A - \lambda_2 I\): \[ (A - \lambda_1 I)v_1 = (A - I)v_1 = \begin{bmatrix} -.1 & .4 \\ .1 & -.4 \end{bmatrix} v_1 = 0 \implies v_1 = \begin{bmatrix} 4 \\ 1 \end{bmatrix}; \] \[ (A - \lambda_2 I)v_2 = (A - .5I)v_2 = \begin{bmatrix} .4 & .4 \\ .1 & .1 \end{bmatrix} v_2 = 0 \implies v_2 = \begin{bmatrix} 1 \\ -1 \end{bmatrix}. \] (b) We have \[ \begin{bmatrix} 3 \\ 2 \end{bmatrix} = \begin{bmatrix} 4 \\ 1 \end{bmatrix} - \begin{bmatrix} 1 \\ -1 \end{bmatrix} = v_1 - v_2, \] so \[ A^k \begin{bmatrix} 3 \\ 2 \end{bmatrix} = A^k v_1 - A^k v_2 = v_1 - (.5)^k v_2. \] Since \((.5)^k\) goes to 0 as \(k\) goes to infinity, the limiting value of \(A^k \begin{bmatrix} 3 \\ 2 \end{bmatrix}\) is \(v_1 = \begin{bmatrix} 4 \\ 1 \end{bmatrix}\). Another argument: the steady state eigenvector of \(A\) is \(\begin{bmatrix} 4 \\ 1 \end{bmatrix}\), so the limit of \(A^k\) as \(k\) goes to infinity is the Markov matrix whose both columns are multiples of \(\begin{bmatrix} 4 \\ 1 \end{bmatrix}\), i.e. \[ A^\infty = \begin{bmatrix} .8 & .8 \\ .2 & .2 \end{bmatrix}, \] and the limiting value of \(A^k \begin{bmatrix} 3 \\ 2 \end{bmatrix}\) is \[ A^\infty \begin{bmatrix} 3 \\ 2 \end{bmatrix} = \begin{bmatrix} 4 \\ 1 \end{bmatrix}. \] (c) Matrices \(A\) and \(B\) are similar if there exists an invertible matrix \(M\) such that \(A = M^{-1}BM\). Equivalently, \(A\) and \(B\) are similar if their Jordan form is the same. The matrix \(A^T\) has the same eigenvalues \(\lambda_1 = 1\) and \(\lambda_2 = .5\) as \(A\), so both are similar to the same Jordan matrix \[ J = \begin{bmatrix} 1 & \\ & .5 \end{bmatrix}. \] Thus \(A\) is similar to \(A^T\).Keywords: Markov matrix, eigenvalues, eigenvectors, steady state, similar matrices