HCM city University of Technology

Exercises and Problems in Linear Algebra

Phan Thi Khanh Van
HCM city University of Technology Exercises and Problems in Linear Algebra

0.1 Complex numbers

0.1.1 Summary
z = a + bi (rectangular form)= r(cos ϕ + i sin ϕ) (polar form)= r.eiϕ (exponential
form)
r = kzk - modulus ofs z; ϕ = arg(z) - argument of z.
z1 = r1 (cos ϕ1 + i sin ϕ1 ), z2 = r1 (cos ϕ1 + i sin ϕ1 )
• Distance between z1 an z2 : d(z1 , z2 ) = kz1 − z2 k
• z1 .z2 = r1 r2 (cos(ϕ1 + ϕ2 ) + i sin(ϕ1 + ϕ2 ))
z1 r1
• z2
= r2
(cos(ϕ1 − ϕ2 ) + i sin(ϕ1 − ϕ2 ))
• z n = rn (cos nϕ + i sin nϕ)
√ √
• n z = n r(cos ϕ+k2π
n
+ i sin ϕ+k2π
n
), k = 0, (n − 1)
Fundamental theorem of algebra: Every non-zero, single-variable, degree n poly-
nomial with complex coefficients has, counted with multiplicity, exactly n complex
roots.
Corollary: P (z) is a polynomial with real coefficients. If z0 is a root of P (z), then z¯0 is
also a root of P (z).

0.1.2 Exercises
√
(1 − 3i)100
1. Evaluate: z =
(1 − i)120
p5
√
2. Evaluate 3−i
3. Solve the equation z 4 + z 3 + 6z 2 − 14z + 20 = 0, given that z = 1 + i is 1 root of
the equation.
√
(− 3 + i)108
4. Find the modulus of z =
(1 − i)1 20
1+i n
5. Find n such that z = ( √ ) is
1 + 3i
• an imaginary number.
• real number.
6. Find the set of all z satisfying
a) |z − 1| = |z + 1 + i|
b) |z + 2 − i| = 5
c) |z + 1 + i| + |z − 2 − 3i| = 4
d) |z − 1 − i| − |z + 2 − i| = 3
e) z = ea+2i , a ∈ R
f ) z = e2+ai , a ∈ R
(
|z − 1| = 3
g) that z is the solution of
|z + i| = |1 − i − z|

Instructor: Phan Thi Khanh Van 2

HCM city University of Technology Exercises and Problems in Linear Algebra

0.2 Matrix, system of linear equations, determinant

0.2.1 Matrix
Elementary row (column) operations of a matrix

1. (Interchange) Interchange two rows (columns): ri ↔ rj

2. (Scaling) Multiply all entries in a row (column) by a nonzero constant: ri →

α.ri , α 6= 0

3. (Replacement) Replace one row (column) by the sum of itself and a multiple of
another row (column) ri → ri + α.rj , ∀α

Row equivalent: Two matrices A and B are called row equivalent: A ∼ B (A is row
equivalent to B) if B can be obtained from A after a finite number of elementary row
operations.
elementary operations
Rank of matrices: Given that A ∈ Rm×n . A −−−−−−−−−−−−→ echelon formE. r(A) =
number of nonzero rows of E.
• r(A) = r(AT )

• r(A) ≤ min{m, n}
Inverse of a matrix: A.A−1 = A−1 .A = In
• (A−1 )−1 = A

• (AT )−1 = (A−1 )T

• (AB)−1 = B −1 A−1
Row elementary operations
Method to find A−1 :[A|In ] −−−−−−−−−−−−−−→ [In |A−1 ]

0.2.2 Systems of linear equations

Gaussian elimination method: A → echelon form A1 , we have the equivalent sys-
tem A1 .X = b1 .
Back substitution.
Croneker -Capelli theorem: Consider the linear system A.X = b, A ∈ Rm×n . If:
• r(A) < r(A|b): there is no solution.

• r(A) = r(A|b) = n: the solution is unique.

• r(A) = r(A|b) = r < n: there is an infinite number of solutions

Homogeneous system of linear equations b = 0
• A.X = 0 always has at least trivial solution X = (0, 0...0)T : r(A) = r(A|0)

• r(A) = n: The trivial solution X = (0, 0...0) is the unique solution

• r(A) < n: There is an infinite number of solutions

Square system

Instructor: Phan Thi Khanh Van 3

HCM city University of Technology Exercises and Problems in Linear Algebra

• The system has unique solution ⇔ r(A) = n ⇔ det(A) 6= 0

• The system has infinitely many solutions ⇔ det(A) = 0 và rank(A|b) = n

• The system has no solution ⇔ det(A) = 0 và rank(A|b) = n + 1

Homogeneous square system b = 0

• The system has unique solution ⇔ r(A) = n ⇔ det(A) 6= 0,

• The system has infinitely many solutions ⇔ r(A) < n ⇔ det(A) = 0

0.2.3 Determinant
Aij cofactor: Deleting the i−th row and the j−th column of A we obtain B.

Aij = (−1)i+j |B|

Using the i − th row ( j− column) to expand the determinant

det(A) = ai1 Ai1 + ai2 Ai2 + ... + ain Ain = a1j A1j + a2j A2j + ... + anj Anj

Effects of Elementary Row Operations on Determinants

r →αr
i i
1. A −− −−→ B ⇒ |B| = α|A|
ri →ri +αrj
2. A −−−−−−→ B⇒ |B| = |A|
ri ↔rj
3. A −−−→ B⇒ |B| = −|A|

The following statements are equivalent

• A is invertible

• r(A) = n

• A ∼ In

• det(A) 6= 0

0.2.4 Exercises
 
  2 −1
1 m 2
1. Given A = , B = 1 4 . Find A + 2B T , 3AB, 2B T AT
3 1 6
2 1
 
1 2 3
2. Given A =  3 1 1 . Find f (A) given that f (x) = 2x2 + 6x − 2.
−2 1 −2

3. A store sells commodities c1 , c2 , c3 in two its branches B1 , B2 . The quantities of

commodities sold in B1 , B2 in a week are given in Table 1, the individual prices
of commodities are given in Table 2, the costs to the store are given in Table 3.
Find the store’s profit for a week

Instructor: Phan Thi Khanh Van 4

HCM city University of Technology Exercises and Problems in Linear Algebra

4. (Markov chain) A machine can be either working or broken down on any given
day. If it is working, it will break down in the next day with probability 5%, and
will continue working with probability 95%. If it breaks down on a given day, it
will be repaired and be working in the next day with probability 80%, and will
continue to be broken down with probability 20%. Given that the probability
that the machine works today is 80% (then, the probability that the machine
breaks down is 20%). Find the probability that the machine will work the day
after tomorrow? after 1 week.

5. ( Leslie model) A population of rabbits raised in a research laboratory has the

characteristics listed below.
(a) Half of the rabbits survive their first year. Of those, half survive their second
year. The maximum life span is 3 years.
(b) During the first year, the rabbits produce no offspring. The average num-
ber of offspring is 6 during the second year and 8 during the third year. The
laboratory population now consists of 24 rabbits in the first age class, 24 in the
second, and 20 in the third. How many rabbits will be in each age class in 1
year, 2 years?

6. (Markov chain) In a city with 1000 householders there are 3 supermarkets A, B

and C. At this month, there are 200, 500 and 300 householders that go to the
supermarkets A, B and C, respectively. After each month, there are 10% of
customers of A change to B, 10% of those change to C; 7% of customers of
B change to A, 3% of those change to C; 8.3% of customers of C change to A,
6.7% of those change to B. Find the numbers of customers of each supermarket
after 1 month, 2 months.

7. (Markov chain) Two competing companies offer satellite  television service

 to a
0.7 0.15 0.15
city with 100 000 households. The transition matrix is 0.2 0.8 0.15 (see the
0.1 0.05 0.7
figure below for the changes in satellite subscriptions each year). Company A
now has 10 000 subscribers and Company B has 15 000 subscribers. How many
subscribers will each company have after 1 year? 2 years?

   
1 −3 2 −1
8. Given A = ,B = . Find X : AX + XB = 2A + 3B T
3 6 1 4

Instructor: Phan Thi Khanh Van 5

HCM city University of Technology Exercises and Problems in Linear Algebra

 
1 −2
9. Find m such that A = is invertible. With m found, find A−1 .
2 m
   
1 −3 −2 3 1 −1
10. Given that A = 3 6 2 , B = 5 3 1 .
4 2 −1 3 1 5
T
Find X : (A + B )X = 3X + 2A − 4B
   
1 −3 −2 3 1 −1
11. Given that A = 3 6 2 , B = 5 3 1 .
4 2 −1 3 1 5
Find X : XA + B T = 3X + 2A

12. Find the rank of the following matrices:

 
1 −3 −2
a) A = 3 6 2
4 2 −1
 
2 1 3
b) A = 1 1 1
0 m 1
 
1 −3 −2 1 1
2 1 3 6 2
c) A = 4 2 −1 3 1


4 −5 −1 8 4
 
1 1 1
d) A = 2 m −2
3 5 4
 
m 1 1
e) A =  1 m 1 
1 1 m
 
1 −2 −1 2
f ) A = 2 −4 m 2 
1 10 −6 m

13. Application in cryptography To encode a message, we convert the text message

into a stream of numerals by associating each letter with its position in the
alphabet: A is 1, B is 2, . . . , Z is 26, and assigning the number 27 to a space
between two words. After that, we break the enumerated message above into a
sequence of 3 × 1 column vectors and place them into a matrix A. Multiplying
this matrix A by the Encoder K (the key),we obtain  the encrypted message

1 1 3 67 67 57
B = KA. Decode the message, where K = 1 2 3, B = 82 94 60.
1 1 2 52 55 46


 2x1 + x2 − x3 + x4 = 3

x − x + x + 4x = −1
1 2 3 4
14. Solve the system


 4x1 − x2 + x3 + 9x4 = 1
x1 + 2x2 − 2x3 − 3x4 = 4


Instructor: Phan Thi Khanh Van 6

HCM city University of Technology Exercises and Problems in Linear Algebra


2x1 − x2 − x3 + x4 = 0

15. Solve the system x1 + 2x2 − x3 − 2x5 = 0

7x1 − x2 − 4x3 + x4 + x5 = 0



mx1 − x2 + 2x3 = 3

16. Find m such that x1 − x2 − x3 = −2 has unique solution

3x1 − x2 − 4x3 = 1



x1 + x2 − 2x3 = 0

17. Find m such that mx1 + x2 + x3 = 0 has nontrivial solution

2x1 + 3x2 + mx3 = 0


18. Find the currents I1 , I2 , I3

19. Find the currents I1 , I2 , I3

20. Find the currents I1 , I2 , I3 , I4 , I5 and I6 in the following electrical network:

Instructor: Phan Thi Khanh Van 7

HCM city University of Technology Exercises and Problems in Linear Algebra

21. The figure shows the flow of traffic of a network. The diagram indicates the av-
erage number of vehicles per hour entering and leaving the intersections and
the arrows indicate the direction of traffic flows. All streets are one-way. Setup
and solve a system of linear equations to find the possible flows xi

22. Setup and solve a system of linear equations to find the possible flows xi

23. Setup and solve a system of linear equations to find the possible flows xi

24. (Input-output Leontief model) Consider a very simple economy that runs on 3
different types of output: raw materials, services, and manufacturing. Raw ma-
terials include the output of many different industries, agriculture and mining
to name two. Services include retailing, advertising, transportation, etc. The

Instructor: Phan Thi Khanh Van 8

HCM city University of Technology Exercises and Problems in Linear Algebra

raw materials industry needs some of the output from the other two industries
to do its job. For example, it needs trucking to get its goods to market, and it
uses some manufactured goods (machines.) The raw materials industry even
needs some of its own output to produce its own output – iron ore to make the
steel to build the rails that carry ore from the mines, for example. Each indus-
try requires some amount of output from each of the three to do its job. All of
these requirements can be summarized in the following table:

Industry Raw materials Services Manufacturing

Raw materials 0.02 0.04 0.04
The numbers in
Services 0.05 0.03 0.01
Manufacturing 0.2 0.01 0.1
the table tell how much output from each industry a given industry requires
in order to produce one dollar of its own output. For example, to provide 1$
worth of service, the service sector requires 0.05$ worth of raw materials, 0.03$
worth of services, and 0.01$ worth of manufactured goods. The demand
matrix D tells how much ( in billions of dollars) of each type of output is
 demanded
  by consumersand others outside the economy
400 0.02 0.04 0.04
D = 200,A = 0.05 0.03 0.01. Let X denote the production matrix. It
600 0.2 0.01 0.1
represents the amounts (in billions of dollars of value) produced by each of
the three industries. Then, we have X = AX + D. Find X.

25. (Market equilibrium) Find the equilibrium price, given the supply function:
Qs = −2 + 3p, the demand function: Qd = 4 − 5p.

26. (Multi-market equilibrium) Find the equilibrium prices in the market of 2 type
of commodities, given the supply and demand functions for:
Commodity 1: Qs1 = −2 + 3p1 ; Qd1 = 10 − 2p1 + p2 .
Commodity 2: Qs2 = −1 + 2p2 ; Qd2 = 15 + p1 − p2 .

27. Find det(A)

 
1 1 1
a) A = 2 m
 −2
3 5 4
 
1 2 3 4
2 5 7 9
b) A = 
1 3

5 6
1 3 5 7
 
1 1 1 m
 2 −2 −2 4 
c) A5 where A = m 5

4 1
1 2 3 4
 
1 1 1 x 

−2 −2 −2 x2 
28. Find the degree of:P (x) = 
3 5 4 x3 
1 2 3 x4 

Instructor: Phan Thi Khanh Van 9

HCM city University of Technology Exercises and Problems in Linear Algebra

 
1 2 m
29. Find m such that A is invertible: A = 4 m 3 
3 −5 4

30. Given A, B ∈ M3 , |A| = 3, |B| = 4. Evaluate:

a) |2AB|
b) |3(AB)−1 |
c) |3A−1 B 2019 |
d) |P3A B −1 |

0.3 Vector spaces

0.3.1 Summary
1. Linear independence
M = {e1 , e2 ...en } is linearly independent (LI) ⇔ ∀(α1 , α2 , ..., αn ) : α1 e1 + α2 e2 +
... + αn en = 0 ⇒ α1 = α2 = ... = αn = 0.

LI ⇔ rank(M ) = n

2. Linear dependence
M = {e1 , e2 ...en } is linearly dependent (LD) ⇔ ∃(α1 , α2 , ..., αn ) 6= (0, 0..0) :
α1 e1 + α2 e2 + ... + αn en = 0

LD ⇔ rank(M ) < n

3. Rank of a vector set

Rank of a vector set M = number of vectors of any maximal independent subset
of M .
rank(M ) = rank(col(M )) = rank(row(M ))
Theorem:

• x is a linear combination of M ⇔ r(M ) = r(M, x)

• x isn’t a linear combination of M ⇔ r(M, x) = r(M ) + 1

4. Spanning set
M = {e1 , e2 ...en } - spanning set of V ⇔ ∀u ∈ V, ∃(α1 , α2 , ..., αn ) : u = α1 e1 +
α2 e2 + ... + αn en

5. Basis

Basis= Spanning set + LI

Dimension dim(V ) = number of vectors in a basis.

Theorem:

• M - spanning set ⇔ rank(M ) = dim(V )

• M - basis ⇔ rank(M ) = dim(V ) = number of vectors of M .

Instructor: Phan Thi Khanh Van 10

HCM city University of Technology Exercises and Problems in Linear Algebra

6. Coordinates  
α1
 α2 
E is a basis of V . 
 ...  = [x]E if x = α1 e1 + α2 e2 + ... + αn en .


αn
In Rn :

• [x]E = E −1 [x], hay [x] = E[x]E

• Let E and F be 2 bases: [x] = E[x]E = F [x]F ⇒ [x]F = F −1 E.[x]E
TE→F = F −1 E - change of basis matrix from E to F .

7. Subspace
The span of a vector set
U = span{e1 , e2 ...em } = {x ∈ V |x = α1 e1 + .. + αm em }
(E = {e1 , e2 ...en } is a spanning set of U )
Nullspace of a homogeneous system
null(A) = {x ∈ Rn |A.x = 0}

dim(null(A)) = n − r(A)

Basis of null(A) is a set of n − r(A) independent solutions of the system.

0.3.2 Exercises
LI, LD

Are the following sets LI or LD:

1. In R3 given M = {(1, 2 − 1), (3, 1, 4), (−2, 0, 1), (1, 2, 3)}

2. In P2 (x) given M = {2x2 + 2x + 3, x − 4, x + 1}

      
1 2 1 1 2 3 4 7
3. In M2 (R) given M = { , , , }
−3 4 2 4 −1 2 −7 10

4. Find m such that 2x2 + x + m, x2 − 2x − m, x + 1 is LI.

5. Find m such that (−2, 1, 3), (2, m, 1)(1, 4, 0), (−2, 2, 1) is LI.

In a vector space V given a LI set {x, y, z}. Are the following set LI or LD:

1. M = {x + 2y, x − 3y, x + y}

2. N = {x + y + z, x − y − z, 2y − z}

3. P = {x + y + 3z, 4x − y + 2z, 2x − 3y − 4z

4. R = {x + y + 2z, x + y + z, 2x, 4y − z}

Instructor: Phan Thi Khanh Van 11

HCM city University of Technology Exercises and Problems in Linear Algebra

Spanning set, basis

1. Is M = {(1, 2 − 1), (3, 1, 4), (−2, 0, 1), (1, 2, 3)} a spanning set of R3 ?

2. Is M = {x2 − 2x + 3, x − 4, 2x + 1} a basis of P2 (R)?

3. In V given a spanning set {x, y, z}. Is M = {x+2y, x−3y, x+y, x+3y} a spanning
set?

4. In V given a basis {x, y, z}. Is M = {x+2y, x−3y, x+y +z, x+3y +z} a spanning
set?

5. Let {x, y, z} be a basis of V . Is M = {x + 2y, x + y − z, x + 3y + z} a basis of V ?

6. Let{x, y, z} be a spanning set of V . Is M = {x + 2y − z, x + 2y − z, x + 3y + z} a

spanning set of V ?

7. Let {x, y, z} be a spanning set of V . Is M = {x + 2y − z, 2x + 2y − 3z, x − 2z} a

spanning set of V ?

8. Let {x, y, z} be a spanning set of V . Is M = {x, 2y, 3z} linearly independent?

Coordinates
1. Let E = {(1, 2 − 1), (3, 1, 4), (−2, 0, 1)} be a basis of R3 .

• Given x = (3, 4, 5). Find [x]E

• Given x = (a, b, c). Find [x]E

2. Let E = {x2 + 2x + 1, 2x2 − x − 2, x + 3}, F = {x2 − x + 2, x2 − 1, 2x + 3} be 2

bases of P2 [R]. Given that [x]E = (3, 1, −4)T , find [x]F

Subspaces
1. In R3 given a subspace U = span{(2, 2, −3), (0, 1, 2), (4, 5, −4), (2, 1, −5)}. Find
one basis and the dimension of U .

2. In R3 given V = {x ∈ R3 |3x1 + 5x2 − 1x3 = 0}. Find one basis and the dimension
of V .

3. In R3 given V = {x ∈ R3 |2x1 + 3x2 − 4x3 = 0, x1 − x2 + 5x3 = 0}. Find one basis

and the dimension of V .

4. In R4 given U =< (1, 2, −1, 4), (3, 1, 2, 4), (1, −3, 4, −4), (4, 3, 1, 8), (1, 2, 3, 4) >.
Find one basis and the dimension of U .

5. In R4 given V = {x ∈ R4 |x1 + 2x2 − x3 + x4 = 0, 2x1 − x2 + 3x3 + 2x4 = 0, 4x1 +

3x2 + x3 + 4x4 = 0, x1 − 3x2 + 4x3 + x4 = 0. Find one basis and the dimension of
V.

6. In R3 given U =< (2, 3, −4), (1, 3, 2), (5, 12, 2) >

• Is x = (3, 4, 5) in U ?
• Find m such that x = (1, 3, m) is in U

Instructor: Phan Thi Khanh Van 12

HCM city University of Technology Exercises and Problems in Linear Algebra

7. In R3 given U = {x ∈ R3 |x1 + x2 − x3 = 0}

• Find one basis E and the dimension of U .

• Find m such that x = (1, 3, m) is in U . Find [x]E

0.4 Inner product spaces

0.4.1 Summary
Inner product spaces = Vector space + inner product. p
Inner product: (x, y) = kxk.kyk.cos(d x, y), norm:kxk = (x, x), distance: d(x, y) =
(x, y)
kx − yk, angle between 2 vectors: x , y = arccos
kxk.kyk
d
Orthogonal:

• u⊥v ⇔ (u, v) = 0

• u⊥M, (M = {e1 , ...en }) ⇔ u⊥ei , ∀i

• M ⊥N, (M = {e1 , ...en }, N = {f1 , ...fm }) ⇔ ei ⊥fj ∀i, j

• u⊥ V ⇔ u⊥ spanning set of V

• U ⊥V ⇔ spanning sets of them are orthogonal.

Orthogonal complement :
Let W be a subspace of V , W ⊥ := {x ∈ V |x⊥W } - is called the orthogonal comple-
ment of W in V .
Projection of a vector onto a subspace:
Let W be a subspace of V . Any vector u ∈ V can be represented uniquely as: u = x +
y, x ∈ W, y ∈ W ⊥ . Then, x - orthogonal projection vector of u onto W : x = prW (u), y
- rejection.
Gram - Schmidt process: Construct an orthogonal set from M = {e1 , ...en }

• f1 = e 1

(e2 ,f1 )
• f2 = e2 − (f 1 ,f1 )
f1 (We can multiply f2 by α so that we don’t have the fractional
elements in f2 .)

(e3 ,f1 ) (e3 ,f2 )

• f3 = e 3 − f
(f1 ,f1 ) 1
− f
(f2 ,f2 ) 2

• ...

(en ,f1 ) (en ,fn )

• fn = e n − f
(f1 ,f1 ) 1
− ... − f
(fn ,fn ) n

Orthonormalization: From M = {e1 , ...en } → orthogonal set → orthogonal set: gi =

fi
kfi k
.

Instructor: Phan Thi Khanh Van 13

HCM city University of Technology Exercises and Problems in Linear Algebra

0.4.2 Exercises
1. Let (x, y) = x1 y1 − 2x1 y2 − 2x2 y1 + 5x2 y2 + 7x3 y3 be an inner product of R3 . Find:

a) k(3, −2, 1)k

b) Find a unit vector u that is parallel to (3, −2, 1)
c) (−1, 1,\
2), (2, 3, 4), d((−1, 1, 2), (2, 3, 4))
d) u = (4, 1, 2), v = (1, 3, 4). Find the norm of 3u − 2v.

2. In an inner product space V given 2 vectors x and y satisfying kxk = 2, kyk =

3, x
d , y = π6 . Find d(2x + 3y, x − 4y)

3. In R4 with the dot product, given U =< (1, 2, −1, 4), (3, −1, 4, −2) >, V = (1, 2, m, n) >.
Find m, n such that V ⊥U

4. In R3 with the dot product, given U =< (1, 2, 3), (−3, 4, 2), (1, 12, 14) >. u =
(4, 5, 7)

a) Find one basis and the dimension of U ⊥

b) Find one orthonormal basis of U .
c) Find the orthogonal projection of u onto U .

5. In R4 with the inner product (x, y) = 2x1 y1 + 3x2 y2 + x3 y3 + 4x4 y4 , given U =<
(1, 3, 2, 1), (2, −1, 1, 0) > và z = (3, 2, 11, 16). Find the orthogonal projection of z
onto U .

6. In R3 with the inner product (x, y) = 3x1 y1 + 2x2 y2 + 4x3 y3 , given F = {x ∈

R3 |2x1 − 3x2 + x3 = 0} and u = (2; 1; 1)

a) Find the orthogonal projection of u onto F .

b) Find the distance from u to F .

7. In R3 with the inner product (x, y) = 4x1 y1 + 3x2 y2 − x2 y3 − x3 y2 + 3x3 y3 , given

U =< (2, 1 − 3), (1, 1, 3) > and z = (2, 3, 4).

a) Find m such that (1, 2, m) is in U

b) Find one basis and the dimension of U ⊥ .
c) Find the orthogonal projection of z onto U ⊥ .

8. In R4 with the dot product, given U =< (2, 1, 3, −1), (3, 2, 1, −2) >. Find the
orthogonal projection of U ⊥

9. In R4 with the dot product, given U =< (1, 1, 4, −1), (3, 1, 1, −2) >.

a) Find one basis and the dimension of U ⊥

b) Find the orthogonal projection of z = (11, −7, 7, 7) onto U ⊥
c) Find the orthogonal projection of w = (3, 2, 11, 16) onto U .

10. In R4 with the dot product, given U = {x ∈ R4 |x1 + x2 − 2x3 + x4 = 0, 2x1 + x2 −

3x3 + x4 = 0, 5x1 + 4x2 − 9x3 + 4x4 = 0}.

a) Find one basis and the dimension of U ⊥

Instructor: Phan Thi Khanh Van 14

HCM city University of Technology Exercises and Problems in Linear Algebra

b) Find one orthonormal basis of U .

11. In R4 with the inner product (x, y) = x1 y1 + 2x1 y2 + 2x2 y1 + 5x2 y2 + x3 y3 + 2x4 y4 ,
given U =< (1, 2, 1, 2), (−3, −1, 2, −2), (1, −3, −4, −2) >.

a) Find one basis and the dimension of U ⊥

b) Find the orthogonal projection of z = (4, 2, 3, −1) onto U .

0.5 Linear transformation

0.5.1 Summary
• Linear transformation: f : U → V : f (ax + by) = af (x) + bf (y)

• Transformation matrix: Let f : Rm → Rn be a linear transformation.

E is a basis of Rm , F is a basis of Rn . AEF - transformation matrix of f in the
bases E, F :

AEF = F −1 f (E), [f (x)]F = AEF [x]E

• Transformation matrix in the standard bases: Let f : Rm → Rn be a linear

transformation. A - Transformation matrix of f in the standard bases of Rm
and Rn :

[f (x)] = A[x]

Let E = {e1 , e2 , ..., em } be a basis of Rm , f : Rm → Rn be a linear transformation.

Given f (e1 ), f (e2 ), ...f (em ).

A = f (E).E −1

• Transformation matrix of f : Rn → Rn in E:

AE = AEE = E −1 f (E)

• Kernel of the linear transformation: Kerf = {x ∈ U |f (x) = 0}

• Image of the linear transformation: Imf = {y ∈ V |∃x ∈ U : y = f (x)}.

U = span{e1 , e2 ...em } ⇒ Imf = span{f (e1 ), f (e2 )...f (em )}

(f (E) is a spanning set of Imf )

Instructor: Phan Thi Khanh Van 15

HCM city University of Technology Exercises and Problems in Linear Algebra

0.5.2 Exercises
1. Given a linear transformation f : R3 → R2 , f (x1 , x2 , x3 ) = (2x1 + x2 − x3 , x1 +
x2 + x3 ).
a) Find f (3, 2, 4)
b) Find one basis, the dimension of Kerf .
c) Find one basis, the dimension of Imf .
2. Let f : R3 → R3 : f (x1 , x2 , x3 ) = (2x1 + x2 − x3 , x1 + x2 + x3 , x1 − 2x3 ) be a linear
transformation
a) Find AEF in 2 bases E = {(1, 1, 1), (1, 1, 2), (1, 0, 0)}, F = {(1, 1, 1), (1, 0, 1), (0, 0, 1)}.
b) Fin AE in the basis E
3. Let f : R3 → R3 be a linear transformation satisfying f (1; 1; 2) = (−1; 2; 1), f (1; 1; 5) =
(2; 2; 3), f (3; 2; 8) = (1; 4; 4)
a) Find f (3; 4; 5)
b) Find f (x1 ; x2 ; x3 )
c) Find one basis, the dimension of Kerf .
d) Find one basis, the dimension of Imf .
4. Let f : R3 → R3 be a linear transformation satisfying f (u1 ) = f (2; 1; 4) =
(1; 2; −1), f (u2 ) = f (1; 1; 5) = (2; 1; 3), f (u3 ) = f (3; 2; 8) = (4; 1; 2).
Find AE of f in the basis E = {(1; 2; 1), (2; 1; 1), (1; 1; 1)}

 matrix of f : R3 → R3 in the basis E = {(1, 1, 0), (0, 1, 2), (0, 3, 1)}

5. Let the transformation

−1 2 3
be AE =  1 1 0. Find one basis, the dimension of the kernel and image of
0 3 3
f.
6. Let f : R3 → R3 be a linear transformation
  with the matrix in the basis E =
1 2 −1
{(1, 1, 1), (1, 1, 2), (1, 2, 1)}: AE = 4 3 0 .
3 1 1
a) Find f (3, −4, 5)
b) Find f (x1 , x2 , x3 )
c) Find the transformation matrix AF of f in the basis F = {(1, 1, 1), (1, 1, 2), (−1, 5, 3)}.
7. In the plane R2 given a figure with the vertices A(3, 3), B(1, 4), C(−3, 3), D(−2, 1), E(2, 1)

Instructor: Phan Thi Khanh Van 16

HCM city University of Technology Exercises and Problems in Linear Algebra

How is this figure mapped if we use

πo
a) Rotation transformation f1 about the origin 0 counterclockwise by 3
. Find
the dimension, one basis for Imf1 , Kerf1 .
b) Reflection transformation f2 about the line y = 2x. Find the dimension, one
basis for Imf2 , Kerf2 .
c) the rotation transformation f1 followed by the reflection transformation f2 .
Find the dimension, one basis for the kernel and image of this composite
transformation.

8. In R3 find the image f (x; y; z), where f is the rotation transformation about the
z axis π4 clockwise from the positive direction of z axis. Find Imf, Kerf .

9. In R3 find the image f (x; y; z), where f is the projection transformation onto
the plane x + 2y − 3z = 0. Find Imf, Kerf .

0.6 Eigenvalues, eigenvectors, diagonalization

0.6.1 Summary
• λ is one eigenvalue of a square matrix A ⇔ ∃x 6= 0 : Ax = λx. The column
vector x is an eigenvector of A

λ is an eigenvalue ofA ⇔ |A − λI| = 0

• P (λ) = |A − λI|- characteristic polynomial.

If A ∈ M2 : P (λ) = λ2 − trace(A)λ + det(A)
If A ∈ M3 : P (λ) = −λ3 + trace(A)λ2 − (A11 + A22 + A33 )λ + det(A)
Solve P (λ) = 0 we obtain the eigenvalues λ1 , λ2 ... where the algebraic multi-
plicity of λi is the multiplicity of the root in the equation.

• The set of all eigenvectors x associated with λi : Ax = λi x - solution of (A −

λu I)x = 0 is called the eigenspace Eλi associated with λi . dim(Eλi ) is called the
geometric multiplicity of λi .

• A is called diagonalizable if A = P DP −1 , where D-diagonal, P - invertible.

A is invertible ⇔ ∀λi : algebraic multiplicity = geometric multiplicity

If A is diagonalizable, then D = diag(λ1 , λ2 , ..., λn ) and P = [e1 , e2 ..., en ]: the

corresponding eigenvectors.

• Properties
Let λ be one eigenvalue of A with the corresponding eigenvectorx.

1. An one eigenvalue λn with the corresponding eigenvector x.

√
2. If B m = A, m is odd, then B has one eigenvalue m λ with the correspond-
ing eigenvector x.
3. If A is invertible then A−1 has one eigenvalue 1
λ
with the corresponding
eigenvector x.

Instructor: Phan Thi Khanh Van 17

HCM city University of Technology Exercises and Problems in Linear Algebra

If A is diagonalizable A = P DP −1 then:

1. Am = P Dm P −1 , where Dm = diag(λm m m
1 , λ2 , ...λn )
√ √ √
2. B m = A, (m is odd) then B = P D1 P −1 , where D1 = diag( m λ1 , m λ2 , ..., m λn )

0.6.2 Exercises
 
2 2 m
1. Given A = 1 3 2m
2 4 1

a) Find m such that x = (2, −1, −1) is an eigenvector of A.

b) Find m such that λ = 1 is an eigenvalue of A.
c) With m found in b), diagonalize A.
 
10 12 6
2. Diagonalize A = −2 0 2 
2 4 −2
 
1 −7 5
3. Diagonalize A = 0 −7 6 . Find B: B 3 = A
2 −13 10
 
−61 31 31
4. Find B such that B 5 = A = −93 63 31.
−93 31 63
 
5 6 3
5. Given A = −1 0 1 . Find A2016
1 2 −1
 
3 2 −3
6. Given A = 3 4 −3. Find A6102
4 4 −3

7. (Markov chain) A machine can be either working or broken down on any given
day. If it is working, it will break down in the next day with probability 5%, and
will continue working with probability 95%. If it breaks down on a given day, it
will be repaired and be working in the next day with probability 80%, and will
continue to be broken down with probability 20%. Given that the probability
that the machine works today is 80% (then, the probability that the machine
breaks down is 20%). Find the probability that the machine will work after 1
years.

8. ( Leslie model) A population of rabbits raised in a research laboratory has the

characteristics listed below.
(a) Half of the rabbits survive their first year. Of those, half survive their second
year. The maximum life span is 3 years.
(b) During the first year, the rabbits produce no offspring. The average num-
ber of offspring is 6 during the second year and 8 during the third year. The
laboratory population now consists of 24 rabbits in the first age class, 24 in the
second, and 20 in the third. Find the stable age distribution vector.

Instructor: Phan Thi Khanh Van 18

HCM city University of Technology Exercises and Problems in Linear Algebra

9. (Markov chain) In a city with 1000 householders there are 3 supermarkets A, B

and C. At this month, there are 200, 500 and 300 householders that go to the
supermarkets A, B and C, respectively. After each month, there are 10% of cus-
tomers of A change to B, 10% of those change to C; 7% of customers of B
change to A, 3% of those change to C; 8.3% of customers of C change to A,
6.7% of those change to B.

a) Find the stable distribution vector (the numbers of supermarkets’ customers

won’t change years after years)
b) Use the model and the method of diagonalization to predict the numbers
of customers of each supermarket after 3 years.

10. (Markov chain) Two competing companies offer satellite television service to
0.7 0.15 0.15
a city with 100 000 households. The transition matrix is 0.2 0.8 0.15 (see
0.1 0.05 0.7
the figure below for the changes in satellite subscriptions each year). Company
A now has 10 000 subscribers and Company B has 15 000 subscribers.

a) Find the stable distribution vector (the numbers of subscribers won’t change
years after years)
b) Use the model and the method of diagonalization to predict the numbers
of subscribers that each company will have after 50 years.

11. Orthogonally diagonalize the following matrices

 
2 −1 −1
a) A = −1 2 −1
−1 −1 2
 
3 −2 4
b) A = −2 6 2
4 2 3

0.7 Quadratic forms

0.7.1 Summary
Quadratic forms:

Instructor: Phan Thi Khanh Van 19

HCM city University of Technology Exercises and Problems in Linear Algebra

f (X) = a11 x21 + a22 x22 + ... + ann x2n + 2a12 x1 x2 + 2a13 x1 x3 + ... + 2aij xi xj = X T .A.X

where A = (aij ) is symmetric: aij = aji , X = (x1 , x2 , ..., xn )T .

Our goal: Transform f (X) into the canonical form for classification:

f (Y ) = a011 y12 + a022 y22 + ... + a0nn yn2

Orthogonal change of variables:

- Orthogonal matrix

P T = P −1

- Orthogonally diagonalization

A = P DP −1 = P DP T

where P - orthogonal matrix.

With the quadratic form f (X), using Y = P −1 X = P T X (or X = P Y ) we have:
f = X T AX = X T .P DP T .X = (P T X)T .D.(P T X) = Y T DY - canonical form.

To orthogonally diagonalization:

• Find the eigenvalues λ1 , λ2 ..., the corresponding eigenspaces Eλ1 , Eλ2 ...

• Find the orthogonal basis for each (Gram-Schmidt): f1 , f2 , ..., fn .

fi
• Normalize: ei = kfi k
.

• The columns of P are the orthonormal eigenvectors:

P = [e1 , e2 , ..., en ], D = diag(λ1 , λ2 ...λn ).

0.7.2 Exercises
Make an orthogonal change of variables to transform the following quadratic form
into the canonical form (the form without cross products)

1. f = x21 + x22 + x23 + 14x1 x2 + 14x1 x3 + 14x2 x3

2. f = x21 + 3x22 − 3x23 + 4x1 x2 + 2x1 x3 + 8x2 x3

3. f = 3x21 + 3x22 + 3x23 − 2x1 x2 − 2x1 x3 − 2x2 x3

4. f = x21 + x22 + x23 + 4x1 x2 + 4x1 x3 + 4x2 x3

5. f = 3x21 + 5x22 + 3x23 − 2x1 x2 + 2x1 x3 − 2x2 x3

Instructor: Phan Thi Khanh Van 20