MA 110

Lecture 01

Saurav Bhaumik
Department of Mathematics
IIT Bombay

Spring 2025

Saurav Bhaumik, IIT Bombay Linear Algebra: Lecture 01

Course Information

Course Code: MA 110

Course Name: Linear Algebra and Differential Equations

Instructors for the first part (Linear Algebra) are Professors

Sudhir Ghorpade and Saurav Bhaumik.

Instructors for the second part (Differential Equations) are

Professors Ronnie M. Sebastian and Mallesham Kummari.

Saurav Bhaumik, IIT Bombay Linear Algebra: Lecture 01

Aim of the course: linear algebra

The first part of this core course aims at an introduction to

the fundamental concepts of linear algebra, including systems
of linear equations, matrices, linear transformations, vector
spaces, eigenvalues, and eigenvectors.

Saurav Bhaumik, IIT Bombay Linear Algebra: Lecture 01

Description/Syllabus of the Linear Algebra part
Vectors in Rn , linear independence and dependence, linear
span of a set of vectors, vector subspaces of Rn , basis of a
vector subspace. Systems of linear equations, matrices and
Gauss elimination, row space, null space, and column space,
rank of a matrix. Determinants and rank of a matrix in terms
of determinants. Abstract vector spaces, linear
transformations, matrix of a linear transformation, change of
basis and similarity, rank-nullity theorem. Inner product
spaces, Gram-Schmidt process, orthonormal bases, projections
and least squares approximation. Eigenvalues and
eigenvectors, characteristic polynomials, eigenvalues of special
matrices (orthogonal, unitary, hermitian, symmetric,
skew-symmetric, normal), algebraic and geometric multiplicity,
diagonalization by similarity transformations, spectral theorem
for real symmetric matrices, application to quadratic forms.
Saurav Bhaumik, IIT Bombay Linear Algebra: Lecture 01
Basic Information (contd..)

Grading Policy
There will be two common quizzes (scheduled on 22 January
and 12 February 2025) and one final exam.
The allotted 50 marks will be split as follows:

Common Quiz 1 : 10 marks

Common Quiz 2 : 10 marks
Final exam : 30 marks

Attendance Policy
Attendance in lectures and tutorials is COMPULSORY.
Students who do not meet 80% attendance will be
awarded the DX grade.

Saurav Bhaumik, IIT Bombay Linear Algebra: Lecture 01

Introduction to linear Algebra

Linear algebra deals with vectors and matrices and, more

generally, with vector spaces and linear transformations.

Linear algebra is central to almost all areas of mathematics.

For example, in geometry, basic objects such as lines, planes
and rotations are represented in terms of linear algebra.
Mathematical analysis applies linear algebra to function spaces.

Linear algebra is also used in most sciences and fields of

engineering, because it allows modelling many natural
phenomena, and computing efficiently with such models.

Saurav Bhaumik, IIT Bombay Linear Algebra: Lecture 01

Introduction to linear algebra contd.

Linear algebra provides a vital arena where the interaction of

Mathematics and machine computation is seen.
Many of the problems studied in Linear Algebra are amenable
to systematic and even algorithmic solutions, and this makes
them implementable on computers.
Numerous Applications within and outside Mathematics. For
example, Google page rank algorithm is based on notions and
results from Linear Algebra.

Saurav Bhaumik, IIT Bombay Linear Algebra: Lecture 01

Notation
N := {1, 2, 3, . . .}
Z := {. . . , −3, −2, −1, 0, 1, 2, 3, . . .}
R := the set of all real numbers
For n ∈ N, let us consider the Euclidean space

Rn := {(x1 , . . . , xn ) : xj ∈ R for j = 1, . . . , n}.

We let 0 := (0, . . . , 0). Also, for x := (x1 , . . . , xn ) and

y := (y1 , . . . , yn ) in Rn , and for α ∈ R, we define

(sum) x + y := (x1 + y1 , . . . , xn + yn ) ∈ Rn ,
(scalar multiple) α x := (α x1 , . . . , α xn ) ∈ Rn ,
(scalar product) x · y := x1 y1 + · · · + xn yn ∈ R.

Saurav Bhaumik, IIT Bombay Linear Algebra: Lecture 01

Matrices
Let m, n ∈ N. An m × n matrix A with real entries is a
rectangular array of real numbers arranged in m rows and n
columns, written as follows:
 
a11 · · · a1k · · · a1n
 .. .. .. 
 . . . 
 
A :=  aj1 · · · ajk · · · ajn 

 = [ajk ],
 .. .. .. 
 . . . 
am1 · · · amk ··· amn
where ajk ∈ R is called the (j, k)th entry of A for
j = 1, . . . , m and k = 1, . . . , n.
Let Rm×n denote the set of all m × n matrices with real
entries. If A := [ajk ] and B := [bjk ] are in Rm×n , then we say
A = B ⇐⇒ ajk = bjk for all j = 1, . . . , m and k = 1, . . . , n.
Saurav Bhaumik, IIT Bombay Linear Algebra: Lecture 01
Let 0 ≤ r < m, 0 ≤ s < n. By deleting r rows and s columns
from A, we obtain an (m − r )×(n − s) submatrix of A.
An n × n matrix, that is, an element of Rn×n , is called a
square matrix of size n.
A square matrix A = [ajk ] is called symmetric if
ajk = akj for all j, k.
A square matrix A = [ajk ] is called skew-symmetric if
ajk = −akj for all j, k.
A square matrix A = [ajk ] is called a diagonal matrix if
ajk = 0 for all j ̸= k.
A diagonal matrix A = [ajk ] is called a scalar matrix if
all diagonal entries of A are equal.
Two important scalar matrices are the identity matrix I in
which all diagonal elements are equal to 1, and the zero
matrix O in which all diagonal elements are equal to 0.
Saurav Bhaumik, IIT Bombay Linear Algebra: Lecture 01
A square matrix A = [ajk ] is called upper triangular if ajk = 0
for all j > k, and lower triangular if ajk = 0 for all j < k.
Note: A matrix A is upper triangular as well as lower
triangular if and only if A is a diagonal matrix.
Examples
 
1 2 3
The matrix 2
 4 5 is symmetric, while the matrix
3 5 6
 
0 2 3
−2 0 5 is skew-symmetric.
−3 −5 0

Note: Every diagonal entry of a skew-symmetric matrix is 0

since ajj = −ajj =⇒ ajj = 0 for j = 1, . . . , n.

Saurav Bhaumik, IIT Bombay Linear Algebra: Lecture 01

Examples
   
2 0 0 2 0 0
The matrix 0 3 0 is diagonal, while 0 2 0 is a scalar
0 0 4 0 0 2
matrix.
 
2 1 −1
The matrix 0 3 1  is upper triangular,
0 0 4
 
2 0 0
while the matrix  1 3 0 is lower triangular.
−1 1 4

Saurav Bhaumik, IIT Bombay Linear Algebra: Lecture 01

A row vector a of length n is a matrix with only one row
consisting of n real numbers; it is written as follows:
 
a = a1 · · · ak · · · an ,

where ak ∈ R for k = 1, . . . , n. Here a ∈ R1×n .

A column vector b of length n is a matrix with only one
column
 consisting
 of n real numbers; it is written as follows:
b1
 .. 
.
b = bk  , where bk ∈ R for k = 1, . . . , n. Here b ∈ Rn×1 .
 
.
 .. 
bn
When n = 1, we may identify [α] ∈ R1×1 with α ∈ R.

Saurav Bhaumik, IIT Bombay Linear Algebra: Lecture 01

Operations on Matrices
Let m, n ∈ N, and let A := [ajk ] and B := [bjk ] be m × n
matrices. Then the m × n matrix A + B := [ajk + bjk ] is called
the sum of A and B. Also, if α ∈ R, then the m × n matrix
αA := [αajk ] is called the scalar multiple of A by α.
These operations follow the usual rules: A + B = B + A,
(A + B) + C = A + (B + C), which we write as A + B + C,
α(A + B) = αA + αB, (α + β)A = αA + βA and
α(βA) = (αβ)A, which we write as αβA. Also, we write
(−1)A as −A, and A + (−B) as A − B.
The transpose of an m × n matrix A := [ajk ] is the n × m
matrix AT := [akj ] (in which the rows and the columns of A
are interchanged).
Clearly, (AT )T = A, (A + B)T = AT + BT and (αA)T = αAT .
Note: A square matrix A is symmetric ⇐⇒ AT = A.
Saurav Bhaumik, IIT Bombay Linear Algebra: Lecture 01
In particular, the preceding operations can be performed on
row vectors, and also on column vectors since they are
particular types of matrices.
Notice that the sum a1 + a2 of two row vectors a1 and a2 of
the same length follows the parallelogram law, and so does the
sum b1 + b2 of two column vectors b1 and b2 of the same
length. (Note: All vectors ‘originate’ from the zero vector.)
Also, note that the transpose of a row vector is a column
vector, and vice versa.  
b1
 .. 
.
We shall often write a column vector b := bk  ∈ Rn×1 , as
 
.
 .. 
bn
 T
b1 · · · bk · · · bn in order to save space.
Saurav Bhaumik, IIT Bombay Linear Algebra: Lecture 01
Let m, n ∈ N. Let α1 , . . . , αm ∈ R.
If a1 , . . . , am ∈ R1×n , then

α1 a1 + · · · + αm am ∈ R1×n

is called a (finite) linear combination of a1 , . . . , am .

Similarly, if b1 , . . . , bm ∈ Rn×1 , then

α1 b1 + · · · + αm bm ∈ Rn×1

is a (finite) linear combination of b1 , . . . , bm .

Saurav Bhaumik, IIT Bombay Linear Algebra: Lecture 01

In particular, for k = 1, . . . , n, consider the column vector
 T
ek := 0 · · · 1 · · · 0 ∈ Rn×1 , where the kth entry is 1
and all other entries are 0.
 T
If b = b1 · · · bk · · · bn is any column vector of length
n, then it follows that b = b1 e1 + · · · + bk ek + · · · + bn en ,
which is a linear combination of e1 , . . . , en . The vectors
e1 , . . . , en are known as the basic column vectors in Rn×1 .

Let A := [ajk ] ∈ Rm×n .

 
Then aj := aj1 · · · ajn ∈ R1×n is called the jth  row
a1
vector of A for j = 1, . . . , m, and we write A =  ... .
 
am
 T
Also, ck := a1k · · · amk is called the kth column vector
of A for k = 1, . . . , n, and we write A = c1 · · · cn .
Saurav Bhaumik, IIT Bombay Linear Algebra: Lecture 01
Examples
   
2 1 −1 1 0 2
Let A := and B := .
0 3 1 −1 4 1
   
3 1 1 10 5 −5
Then A + B := and 5A = .
−1 7 2 0 15 5
   
The row vectors of A are 2 1 −1 and 0 3 1 .
     
2 1 −1
The column vectors of A are and and .
0 3 1
 
2 0
Also, AT =  1 3.
−1 1

Saurav Bhaumik, IIT Bombay Linear Algebra: Lecture 01

Matrix multiplication
So if m, n, p ∈ N, A := [ajk ] ∈ Rm×n and B := [bjk ] ∈ Rn×p ,
then AB ∈ Rm×p , and for j = 1, . . . , m; k = 1 . . . , p,
n
X
AB = [cjk ], where cjk := aj bk = ajℓ bℓk .
ℓ=1

Note that the (j, k)th entry of AB is a product of the jth row
vector of A with the kth column vector of B as shown below:
  b1k

..
 .
 
  
aj1 · · · a jℓ · · · a jn 
 
 b ℓk


..
  
.
  
bnk
Clearly, the product AB is defined only when the number of
columns of A is equal to the number of rows of B.
Saurav Bhaumik, IIT Bombay Linear Algebra: Lecture 01
Note that AI = A, IA = A, AO = O and OA = O whenever
these products are defined.
Examples  
  1 6 0 2
2 1 −1
(i) Let A := and B := 2 −1 1 −2 .
0 3 1 2×3
2 0 −1 1 3×4
 
2 11 2 1
Then AB = .
8 −3 2 −5 2×4
 
1 0
(ii) In general, AB ̸= BA. For example, if A := and
     0 0 
0 1 0 1 0 0
B := , then AB := , while BA := .
0 0 0 0 0 0
Note that BA = O, while AB = B ̸= O. Since A ̸= I,
we see that the so-called cancellation law does not hold.

Saurav Bhaumik, IIT Bombay Linear Algebra: Lecture 01

Let A := [ajk ] ∈ Rm×n , and let e1 , . . . , en be the basic column
vectors in Rn×1 . Then for k = 1, . . . , n,
 
a1k
 .. 
 . 
A ek =  ajk , which is the kth column of A.
 
 . 
 .. 
amk

Saurav Bhaumik, IIT Bombay Linear Algebra: Lecture 01

Properties of Matrix Multiplication
Consider matrices A, B, C and α ∈ R. Then it is easy to see
that (A + B)C = AC + BC, C(A + B) = CA + CB and
(αA)B = αAB = A(αB), if sums & products are well-defined.
Matrix multiplication also satisfies the associative law:
Proposition
Let m, n, p, q ∈ N. If A ∈ Rm×n , B ∈ Rn×p and C ∈ Rp×q ,
then A(BC) = (AB)C (which we shall write as ABC).

Proof. Let A := [ajk ], B := [bjk ] and C := [cjk ]. Also, let

(AB)C := [αjk ] and A(BC) := [βjk ]. Then
p  n  n p
X 
X X X
αjk = ajℓ bℓi cik = ajℓ bℓi cik = βjk
i=1 ℓ=1 ℓ=1 i=1

for j = 1, . . . , m and k = 1, . . . , q. Hence the result.

Saurav Bhaumik, IIT Bombay Linear Algebra: Lecture 01
Also, the transpose of a product is the product of the
transposes in the reverse order:
Proposition
Let m, n, p ∈ N. If A ∈ Rm×n and B ∈ Rn×p , then
(AB)T = BT AT .

Proof. Let A := [ajk ], B := [bjk ] and AB := [cjk ].

′ ′
Also, let AT := [ajk ], BT := [bjk ] and (AB)T := [cjk′ ]. Then
n
X n
X
cjk = ajℓ bℓk and so cjk′ = ckj = akℓ bℓj
ℓ=1 ℓ=1

for j = 1, . . . , m; k = 1, . . . , p. Suppose BT AT := [djk ]. Then

n
X n
X
′ ′
djk = bjℓ aℓk = bℓj akℓ = cjk′
ℓ=1 ℓ=1

for j = 1, . . . , m; k = 1, . . . , p. Hence the result.

Saurav Bhaumik, IIT Bombay Linear Algebra: Lecture 01
Matrix Multiplication Revisited
Let m, n, p ∈ N, A ∈ Rm×n and B ∈ Rn×p . For j = 1, . . . , m,
let aj := aj1 · · · ajn be the  jth row of A, and for
b1k
 .. 
k = 1, . . . , p, let bk :=  .  be the kth column of B. Then
  bnk
a1
A =  ... , in terms of its rows. Also, B = b1 · · · bp , in
   

am
terms of its columns. Note: AB has m rows and p columns.
Rows of AB
Fix j ∈ {1, . . . , m}, and consider the jth row of AB, namely,
aj B = [aj b1 · · · aj bp ]. For k = 1, . . . , p, the kth entry of the
jth row of AB is aj bk = aj1 b1k + · · · + ajn bnk , where
b1k , . . . , bnk are the kth entries of the n row vectors of B.
Saurav Bhaumik, IIT Bombay Linear Algebra: Lecture 01
Thus we see that for j = 1, . . . , m, the jth row of AB is a
linear combination of the n row vectors of B with coefficients
aj1 , . . . , ajn provided by the jth row of A.
Columns of AB
Fix k ∈ {1, . . . , 
p}, and consider the kth column of AB,
a1 bk
 .. 
namely, Abk =  . . For j = 1, . . . , m, the jth entry of
am bk
the kth column of AB is aj bk = aj1 b1k + · · · + ajn bnk , that is,
b1k aj1 + · · · + bnk ajn , where aj1 , . . . , ajn are the jth entries of
the n columns of A.
Thus we see that for k = 1, . . . , n, the kth column of AB is a
linear combination of the n column vectors of A with
coefficients b1k , . . . , bnk provided by the kth column of B.
These descriptions of the rows and columns of AB are useful.
Saurav Bhaumik, IIT Bombay Linear Algebra: Lecture 01
Example
As we have seen,
 
  1 6 0 2  
2 1 −1  2 11 2 1
2 −1 1 −2 =
 , where
0 3 1 8 −3 2 −5
2 0 −1 1
     
2 11 2 1 = 2 1 6 0 2 + 1 2 −1 1 −2
 
−1 2 0 −1 1 ,
     
8 −3 2 −5 = 0 1 6 0 2 + 3 2 −1 1 −2
 
+1 2 0 −1 1 ,
       
2 2 1 −1
= 1 +2 +2 ,
8 0 3 1
       
11 2 1 −1
= 6 −1 +0 , etc.
−3 0 3 1
.
Saurav Bhaumik, IIT Bombay Linear Algebra: Lecture 01
Linear System
Let m, n ∈ N. A linear system of m equations in the n
unknowns x1 , ..., xn is given by

a11 x1 + a12 x2 + · · · + a1n xn = b1 (1)

a21 x1 + a22 x2 + · · · + a2n xn = b2 (2)
.. .. .. .. ..
. . . . .
am1 x1 + am2 x2 + · · · + amn xn = bm , (m)

where ajk ∈ R for j = 1, . . . , m; k = 1, . . . , n and also bj ∈ R

for j = 1, . . . , m are given.
 T
Let A := [ajk ] ∈ Rm×n , x := x1 · · · xn ∈ Rn×1 and
 T
b := b1 · · · bm ∈ Rm×1 . Using matrix multiplication, we
write the linear system as
Ax = b.
Saurav Bhaumik, IIT Bombay Linear Algebra: Lecture 01