MAT3341: Applied Linear Algebra Final Exam 2018
Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit
MAT3341: Applied Linear Algebra Final Exam
2018
[3] 1. Find all real vectors x and y such that xyT = ┌0(1)
'4
一8'. No justification is required.
[7] 2. You are given the following row reductions involving A and B .
[A
[B
[C |
I]
I]
I] |
┌ 0(1) ' 0 ┌ 0(1) ' ┌' 0(1) '(') 0 ' 0 ' 0 |
0 0 0 0 1 0 0 1 0 0 0 |
|
a) Give the size and rank of each of A and B and C .
b) Give the general left inverse and the general right inverse for A and B and C . You should give your answer as a sum/product of matrices (as we saw in class). No justification is required.
[7] 3. Given that PAQ = L0U0 with the following matrices.
P = ┌0(1) 0(0) '0 1 |
1(0)┐ 0' |
Q = ┌0(0) 0(1) '1 0 |
1(0)┐ 0' |
L0 = ┌ 1一1 ' 2 |
0 1 1 |
0(0)┐ 1' |
U0 = ┌0(1) '0 |
0 2 0 |
1(1)┐ 0' |
a) Give the size and rank of A.
b) Solve Ax = '(┌)'(┐) and Ax = '(┌)'(┐) . Use the method based on this decomposition we saw in class, not some other method. (hint: exactly one of these has no solution)
[3] 4. Let A = '(┌)0(1) 2(0) 0(0)'(┐) . Show that 冷x卜yS is an inner product for R3 .
[6] |
5. Let A = ┌ 1(1) 2(0) '0 2 |
1 0 一1 |
1(0)┐ 0'. |
a) Find a Q0R0 decomposition for A.
b) Find a QR decomposition for A.
c) Find a matrix P such that Px is the projection of x onto the column space of A. You may give P as a matrix product of other matrices you have determined without multiplying them out.
6. A has a QR decomposition with Q = ┌ '1/^2 |
1/^2 一1/^2 0 |
一22^2'(┐) and R = '(┌) |
1一1┐ 1 '. |
Using this decomposition, and not some other method, find a best solution x to Ax ↓ ┌ ┐1(1)
'0'.
7. Consider 扌3, the space of polynomials in the variable t of degree at most 3, and R3, with the following bases.
., '(┌)'(┐)、.
Let T : 扌3 → R3 be the linear map as follows (you don’t need to check that it is linear).
T (p(t)) = ┌┐
'p(2)'
Find the matrix of T with respect to these two bases.
[2] 8. Show that if A is a diagonal matrix, then 卜A卜1 = 卜A卜2 = 卜A卜/ .
[5] 9. For A = -2(1) 1(2)┐ and b = - ┐3(3) , we find that x = - ┐1(1) is the solution to Ax = b. Now we want
to solve A-x- = b where A- = A+∆A, where ∆A is some unknown matrix all of whose entries
are in the range 士0.1. Let ∆x = x- 一 x. In case it is useful to you, A一1 = -2(一) 一(2)┐ .
a) Find an upper bound on 卜∆x卜1 using the methods we saw in class.
b) Using your bound on 卜∆x卜1 , find the best possible ∈ so that the entries of ∆x are all in the range 士∈ .
c) Give the tightest possible range for each of the entries of x- .
[5] 10. Let A = PDP一1 with D = '(┌)一01 2(0) 0(0)'(┐) and P = '(┌)2(1) 3(1) 4(1)'(┐); also P一1 = '(┌) 0(2) 一11 1一2'(┐)
Let A- = A + ∆A where ∆A is some unknown matrix all of whose entries are in the range 士0.3. Using a Theorem from class, describe the approximate location in the complex plane of the eigenvalues of A- . Use the &-norm. Be as detailed as you can in your description. (hint: calculating A will involve a lot of work and will not help)
[5] 11. Let A = -e(1) e 2(一)1 ┐ where e > 0. We wish to choose e so as to make the condition number
of A as small as possible. The campaign to “MAKE EPsILoN SMALL AcAIN!” proposes to do this by choosing e as close as possible to 0, in order to prevent imaginary numbers from getting in.
a) Verify (any correct method you prefer) that A一1 = -2一e 一 1(e)一1 ┐ .
b) Compute the condition number of A, in terms of e. Do this using the 1-norm. You do not need to simplify the formula, it just needs to be a formula in terms of e and not a matrix.
c) Will the strategy proposed by MeSA work? Find the limit as e → 0 of the condition number of A, and decide if making e close to zero would make the condition number small. You do not need to find the value of e that minimizes the condition number.
[5] 12. We want to find a Schur decomposition for the matrix A = '(┌) 5 3一21'(┐) . We want
to find a Schur decomposition. The first step is done for you: Following the notation used in class, we set A1 = A and we find (after some work) the matrix Q1 , and then obtain S1 = Q1(H)A1Q1 , to get:
Q1 = ┌ ' 0 |
4/5 一3/5 0 |
0(0)┐ 1' |
S1 = Q1(H)A1Q1 = ┌0(3) '0 |
1 1 2 |
3(1)┐ 2' |
Finish the decomposition. You should find all matrices explicitly, except for the final P and S: these two you can give as a product of other matrices that are either given in the question or you have already found without explicitly multiplying them out.
[5] 13. Let A = - i(1) |
i 1 |
1(i)┐ . |
a) Find a singular value decomposition for A. As a hint, the eigenvalues of AHA should turn out to be 2 and 4.
b) Find the rank 1 matrix that best approximates A, in the sense of the 2-norm. You may
give your answer in the form of a product of matrices, no need to multiply it out.
2022-07-25