MATH 3300 Final Exam - Sample (Solutions)
Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit
MATH 3300 Final Exam - Sample (Solutions)
1. (4+3+3 points) Let
A = [ ⃗a1
⃗a2 ⃗a3 ] = l 0(5)
3
−8
−2(9)
(a) Explain if {⃗a1 ,⃗a2 ,⃗a3 } is linearly independent.
(b) Explain if {⃗a1 ,⃗a2 } is linearly independent.
(c) Explain if A is invertible.
Solutions:
(a) Reduce A to row echelon form,
A ∼ l−0(5) 「 0
3
− 2
− 9(1)
This show that there are 3 pivots so the rank of A is T = 3. which is the number n of columns of A . Hence the set of columns {⃗a1 ,⃗a2 ,⃗a3 } is linearly independent.
(b) The set {⃗a1 ,⃗a2 } is a part of the set {⃗a1 ,⃗a2 ,⃗a3 } that is linearly independent by (a), it is also linearly independent.
(c) A is invertible since it has full rank 3 (T = n = 3) by the Invertible Matrix Theorem.
2. (10 points) (a) Find A − 1 for
A =「 0
−1(a)
0
− b
(b) Find X in AX = l 2(1) −2(0) 」
Solutions: (a) Apply elementary row operations
「 0 0 1 0 0 1 ~「 0 0 1 0 0 1 ~「 0 0
A − 1 = l 」
「 0 0 1
(b) Let B = 「(l) . Then X = A − 1 B = 「(l) 1 + 2 − .
3. (8+2 points) Given
2
A = 「 −6(1)
−5(4)
− 2
−4(4)
(a) Find the LU factorization of A .
(b) Explain how to use (a) to quickly find |A| with almost no more computation. Solutions: (a)
A = l 21 − 」 l 0(2) 」 l 0(2) −7(4) − 5(2) 」
−6 − 2 4 0 − 14 10 0 0 0
To find L , perform the following operations on the first matrix below, divide the first column by the circled number (pivot) 2,
divide the second column by circled number (pivot) 7,
「(l) 6(1) − 1(7)4 1 ❀ 「(l) − − 2(1) 1 = L
Therefore,
A = LU = l 1(1) 1(0) 0(0) 」 l 0(2) −7(4) − 5(2) 」
(b) |A| = |LU| = |L||U| = 0.
4. (4+3+1+1 points) Let T : R4 −→ R3 be defined by
T \图图( l''「 」地地l)图图) = 「(l) l(」)
(a) Find the matrix A for T under the standard bases and find its reduced row echelon form.
(b) Find a basis for each of the column space C(A), row space Row (A), and null space N(A).
(c) Is the transformation T one-to-one? Justify your answer.
(d) Is the transformation T onto? Justify your answer.
Solutions: (a) A == [ ⃗a1 ⃗a2 ⃗a3 ⃗a4 ] = [ T(⃗e1 ) T(⃗e2 ) T(⃗e3 ) T(⃗e4 ) ] = l 0(1) |
− 1(1) 0 |
0 − 1 |
0(0) 」 − 1 l , |
RREF = l 0(1) 1(0) 「 0 0 |
0 0 1 |
− 1 」 1(1) l |
(b) Using the RREF, a basis for C(A) is {⃗a1 ,⃗a2 ,⃗a3 } ,
a basis for Row (A) is {(1, 0, 0, − 1), (0, 1, 0, − 1), (0, 0, 1, − 1)} ,
and a basis null space N(A) is {(1, 1, 1, 1)} ,
where all vectors are column vectors.
(c) First explanation: the null space N(A) is non-trivial, so T is not one-to-one.
Second explanation: Matrix A is of size m × n = 3 × 4. From the RREF of A, the rank r = 3 < n, therefore T is not one-to-one.
(d) From the RREF of A, the rank r = 3 = m, therefore T is onto.
5. (4+3+3 points) (a) Compute the determinant |A|, where
l 1 − 2 5 2 」
A =
「 2 0 3 5
(b) Use part (a) to justify that A − 1 exists.
(c) Find the following values
|AT A − 1 | =
|A− 1 AT A| =
| − A| =
Solutions: (a) Expand along second row of A, then along the last row of the resulting matrix,
' 1 − 2 2 '
(b) From (a), since |A| 0, A − 1 exists by the Invertible Matrix Theorem.
(c) |AT A − 1 | = |AT ||A− 1 | = |A||A− 1 | = 1
|A− 1 AT A| = |A− 1 ||AT ||A| = |A| = 12
| − A| = ( − 1)4 |A| = 12
6. (2+6+2 points) Let
A = [ ⃗a1
⃗a2 ⃗a3 ] = l1(1) 1(0)
(a) Explain that {⃗a1 , ⃗a2 , ⃗a3 } is a basis for the column space C(A).
(b) Construct an orthogonal basis {⃗v1 , ⃗v2 , ⃗v3 } using {⃗a1 , ⃗a2 , ⃗a3 } .
(c) Find an QR Factorization for A .
Solutions: (a) |A| = 1 0, by invertible matrix theorem, the set of columns {⃗a1 , ⃗a2 , ⃗a3 } is linearly independent, and is therefore a basis for C(A) by definition (of a basis).
(b) We apply GSP (Gram-Schmidt Process using what we called the Beautiful Formula for the projections in class) to the basis in (a), we obtain (omitting the algebra) an orthonormal basis {⃗u1 ,⃗u2 ,⃗u3 } with
⃗u1 = ( , , ),⃗u2 = ( , , ),⃗u3 = (0, , ),
each being a column vector.
(c) Since Q = [ ⃗u1 ⃗u2 ⃗u3 ],
R = QT A = 「(l) [ ⃗a1 l R =「 l 「 |
⃗a2
3 ^3 0 0
1 ^6 1 ^6 |
⃗a3 ] =
|
l ⃗u⃗a1 「 0
and
」
|
7. (10 points) Let
A = l 0(4)
「 1
2(0) 」 = l 0(2) 」
1 , 「 11
(1) Find a least square solution of A⃗x = .
(2) Find the distance from to the column space C(A).
Solutions:
(1) We solve the system of normal equations AT Ax∗ = , where AT A = ] , = 「(l) 1
There is only one least square solution for this problem (though not true in general):
x∗ = [2(1) ]
(2) The distance from to the column space C(A) is the same as the least square error,
i.e.,
∥ − Ax∗ ∥ = 「(l) 1 − 「(l)
[2(1) ] = ^84
8. (10 points) For the quadratic form
Q(⃗x) = 2x1(2) − 6x2(2) + 6x1 x2 ,
find an orthogonal change of variables to transform Q(⃗x) into a new form without cross terms, and write down the new form explicitly.
Solutions: Note that Q(⃗x) = ⃗xT A⃗x with (symmetric) matrix for Q(⃗x) given by
A = [3(2) −6(3) ] ,
which has eigen-values λ 1 = − 7 and λ2 = 3.
A basis for the eigen-space N(A− λ 1 I) is [ −]3(1) , which is normalized as ⃗u1 = [ 0(0) ]. A basis for the eigen-space N(A − λ2 I) is [1(3) ], which is normalized as ⃗u2 = [ 0(0) ].
(Note: for symmetric matrix A with larger size n, one may need to apply Gram-
Schmidt Process to obtain orthonormal bases for some eigen-spaces N(A − λI).)
Hence the orthogonal matrix P =[ ⃗u1 ⃗u2 ] = [ 0(0)
change of variables is ⃗x = P⃗y, which gives
Q(⃗x) = Q(P⃗y) = ⃗yT PT AP⃗y = ⃗yT [ −0(7) 3(0) ] ⃗y = − 7y1(2) + 3y2(2) .
9. (13 points) Find a SVD (Singular Value Decomposition) of
A = l 0(1) 1(1) 」
「−1 1
Solutions: AT A = [0(2) 3(0) ]. Characteristic equation |AT A − λI| = (λ − 3)(λ − 2). Eigen value λ1 = 3, with normalized eigen-vector ⃗v1 = [1(0) ]. Eigen value λ2 = 2, with normalized eigen-vector ⃗v2 = [0(1) ]. V = [ ⃗v1 ⃗v2 ]= [1(0) 0(1) ].
Singular values σ1 =^λ1 =^3, σ2 =^λ2 =^2. Σ = l 」
「 0 0 .
Next ⃗u1 = = 「(l) , ⃗u2 = = 「(l) − , and ⃗u3 can be found by solving ⃗y
from system of equations ⃗u1(T)⃗y = 0, ⃗u2(T)⃗y = 0, or equivalently AT ⃗y = 0, i.e.,
A basis for the solutions space is ⃗y =
l 」 Hence ⃗u3 = = 「 , and U == [ ⃗u1 ⃗u2 The SVD is A = UΣVT , or |
l ⃗u3 ] == 「 |
l 1 」
「 − 1(2) .
|
2022-12-11