Measure Theory and Integration

Ryan Azim Shaikh

February 2025

Ryan Azim Shaikh Measure Theory and Integration

𝜎-algebra

𝜎-Algebra
A 𝜎-algebra on a non-empty set 𝑋 is a class of subsets of 𝑋 that is
closed under countable union and countable intersection.
The empty set ∅ and the full set 𝑋 always belong to a 𝜎-algebra on
𝑋. These can be considered, as per convention, as the empty union
and empty intersection, respectively. Some books add these to the
𝜎-algebra as part of the definition.

Ryan Azim Shaikh Measure Theory and Integration

Borel 𝜎-algebra

Borel 𝜎-algebra
The Borel 𝜎-algebra is the smallest 𝜎-algebra that contains the
topology on a topological space. A set in the Borel 𝜎-algebra is
called a Borel set.
The Borel 𝜎-algebra contains all
Open Sets.
Closed Sets.
Intervals.
𝐺𝛿 sets.
𝐹𝜍 sets.

Ryan Azim Shaikh Measure Theory and Integration

Positive Measure

A positive measure
A non-negative function 𝑚 mapping a 𝜎-algebra ℱ on 𝑋 to the
extended real line ℝ is said to be a positive measure if:
𝑚(∅) = 0.
it is countably additive.

Ryan Azim Shaikh Measure Theory and Integration

Lebesgue Outer Measure

Lebesgue Outer Measure

The Lebesgue outer measure is a set function defined on the
subsets of ℝ as:
∞ ∞
⋆
𝑚 (𝐴) = inf { ∑ 𝑙 (𝐼𝑛 ) ∣ {𝐼𝑛 }𝑛∈ℕ ⊂ ℐ, 𝐼𝑛 ⊇ 𝐴} .
⋃
𝑛=1 𝑛=1

where ℐ is the set of all open bounded non-empty intervals in ℝ.

Ryan Azim Shaikh Measure Theory and Integration

Properties of Outer Measure

Outer measure of an interval

The outer measure of an interval is its length.

Translation Invariance of Outer Measure

The outer measure is translation invariant, i.e., 𝑚⋆ (𝐴) = 𝑚⋆ (𝐴 + 𝑥)
for all 𝐴 ⊆ ℝ and 𝑥 ∈ ℝ.

Countable Subadditivity of Outer Measure

The outer measure is countably subadditive, i.e.,
𝑚⋆ (⋃𝑛=1 𝐴𝑛 ) ≤ ∑𝑛=1 𝑚⋆ (𝐴𝑛 ).
∞ ∞

Ryan Azim Shaikh Measure Theory and Integration

Lebesgue Measurable Set

Lebesque Measurable Set

A subset 𝐴 of ℝ is said to be measurable if:

𝑚⋆ (𝐸) = 𝑚⋆ (𝐸 ∩ 𝐴) + 𝑚⋆ (𝐸 ∩ 𝐴𝐶 ) ∀𝐸 ⊆ ℝ

Ryan Azim Shaikh Measure Theory and Integration

Properties of Lebesque Measurable sets

Any set of outer measure zero is measurable. Every countable

subset of ℝ has outer measure zero and hence is measurable.
The set of Lebesgue measurable sets is a 𝜎-algebra.
Every interval is measurable.
Every Borel set is measurable.

Ryan Azim Shaikh Measure Theory and Integration

Properties of Measurable Sets

Let 𝐸 be a set of real numbers. Then the following are equivalent:

𝐸 is measurable.
For each 𝜖 > 0, there exists an open set 𝒪𝜖 containing 𝐸 such
that 𝑚⋆ (𝒪𝜖 ∼ 𝐸) < 𝜖.
There exists a 𝐺𝛿 set 𝐺 containing 𝐸 such that 𝑚⋆ (𝐺 ∼ 𝐸) = 0.
For each 𝜖 > 0, there exists a closed set 𝒦𝜖 contained in 𝐸
such that 𝑚⋆ (𝐸 ∼ 𝒦𝜖 ) < 𝜖.
There exists an 𝐹𝜍 set 𝐹 contained in 𝐸 such that
𝑚⋆ (𝐸 ∼ 𝐹) = 0.

Ryan Azim Shaikh Measure Theory and Integration

Approximation of Measurable Sets

Theorem
Let 𝐸 be a set with finite outer measure. Then for each 𝜖 > 0,
there exists an integer 𝑛 and a finite disjoint collection {𝐴𝑖 }𝑛𝑖=1 of
open intervals for which if 𝒪 = ⋃𝑖=1 𝐴𝑖 , then:
𝑛

𝑚⋆ (𝐸 ∼ 𝒪) + 𝑚⋆ (𝒪 ∼ 𝐸) < 𝜖

Ryan Azim Shaikh Measure Theory and Integration

Lebesgue Measure

Lebesque measure
The restriction of the Lebesgue outer measure 𝑚⋆ to the 𝜎-algebra
of Lebesgue measurable sets is the Lebesgue measure.

Ryan Azim Shaikh Measure Theory and Integration

Properties of Lebesgue Measure

The Lebesgue measure is countably additive.

The Lebesgue measure is a measure on ℝ that is translation
invariant and assigns length to intervals.
(Continuity of Lebesgue Measure) Given an increasing
sequence of measurable sets {𝐸𝑖 }∞
𝑖=1
, we have:
∞
lim 𝑚(𝐸𝑛 ) = 𝑚 (⋃ 𝐸𝑖 )
𝑛→∞
𝑛=1

Given a decreasing sequence of measurable sets {𝐸𝑖 }∞

𝑖=1
with
𝑚(𝐸1 ) < ∞, we have:
∞
lim 𝑚(𝐸𝑛 ) = 𝑚 (⋂ 𝐸𝑖 )
𝑛→∞
𝑛=1

Ryan Azim Shaikh Measure Theory and Integration

Borel-Cantelli Lemma

Borel-Cantelli Lemma
Let {𝐸𝑖 }𝑖∈ℕ be a countable collection of measurable sets such that
∑𝑖=1 𝑚(𝐸𝑖 ) < ∞. Almost all points in ℝ belong to at most finitely
∞

many of the 𝐸𝑖 ’s.

Ryan Azim Shaikh Measure Theory and Integration

Non-Measurable sets

Proposition
If 𝐸 is a bounded measurable set of real numbers and Λ is a
bounded countably infinite set of real numbers such that the class
of sets {𝜆 + 𝐸 ∣ 𝜆 ∈ Λ} is pairwise disjoint, then 𝑚(𝐸) = 0.

Theorem (Vitali)
Any set 𝐸 of real numbers with positive outer measure contains a
subset that is not measurable.

Ryan Azim Shaikh Measure Theory and Integration

Corollary

Non-Additivity of Outer Measure

There are disjoint sets 𝐴 and 𝐵 of real numbers such that:

𝑚⋆ (𝐴 ∪ 𝐵) < 𝑚⋆ (𝐴) + 𝑚⋆ (𝐵).

Ryan Azim Shaikh Measure Theory and Integration

The Cantor Set

Consider the sequence of closed sets constructed as follows.

𝐶1 = [0, 1], 𝐶𝑛 is formed by subdividing each interval in 𝐶𝑛−1 into
three equal length intervals, with the middle subdivision being
open and deleting the middle subdivision creating two new closed
intervals from each interval in 𝐶𝑛−1 . Thus, 𝐶2 = [0, 1/3] ∪ [2/3, 1],
𝐶3 = [0, 1/9] ∪ [2/9, 1/3] ∪ [2/3, 7/9] ∪ [8/9, 1] and so on. The union
of open intervals deleted on creating 𝐶𝑛+1 from 𝐶𝑛 is defined as 𝑂𝑛 .
The Cantor Set
The Cantor set C is defined as the intersection of the decreasing
sequence of non-empty closed sets 𝐶𝑛 .

Properties of the Cantor Set

The Cantor set C is a closed, uncountable, measurable set of
measure zero.

Ryan Azim Shaikh Measure Theory and Integration

The Cantor-Lebesque function

the Cantor-Lebesgue Function

The Cantor-Lebesgue function 𝜑 is an increasing continuous
function that maps [0, 1] onto [0, 1]. Its derivative exists on the
open set 𝑂 = ⋃𝑛=1 𝑂𝑛 , the complement of the Cantor set in [0, 1],
∞

and 𝜑′ (𝑥) = 0 on 𝑂 with 𝑚(𝑂) = 1.

Ryan Azim Shaikh Measure Theory and Integration

Construction of the Cantor-Lebesgue Function

The Cantor-Lebesgue function 𝜑 is constructed step by step. It is

defined as an increasing function on the complement of the Cantor
set 𝑂 and then extended to the Cantor set C.
On 𝑂, 𝜑 is constant on each of the open intervals in 𝑂𝑘 and takes
2𝑘 −1
the values { 𝑘 , 𝑘 , ⋯ , 𝑘 } starting at the left-most interval (on
1 2
2 2 2
the real line) and going sequentially.
On C, 𝜑(𝑥) = sup{𝜑(𝑡) ∣ 𝑡 ∈ 𝑂 ∩ [0, 𝑥)}.

Ryan Azim Shaikh Measure Theory and Integration

Mapping Properties of the Cantor-Lebesgue Function

Let 𝜑 be the Cantor-Lebesgue function and define the function 𝜓

as 𝜓(𝑥) = 𝜑(𝑥) + 𝑥 for all 𝑥 ∈ [0, 1]. Then 𝜓 is a strictly increasing
continuous function that maps [0, 1] onto [0, 2] and:
Maps the Cantor set onto a measurable set of positive
measure.
Maps a measurable set, a subset of the Cantor set, onto a
nonmeasurable set.

Ryan Azim Shaikh Measure Theory and Integration

Non-Borel Subset of the Cantor Set

Proposition
There is a measurable subset of the Cantor set that is not Borel.

Ryan Azim Shaikh Measure Theory and Integration

Lebesque Measurable Functions

Ryan Azim Shaikh Measure Theory and Integration

Characterization of Measurable Functions

Let 𝑓 be a function with measurable domain 𝐸, then the following

are equivalent:
1 {𝑥 ∈ 𝐸 ∣ 𝑓(𝑥) > 𝑐} is measurable for each 𝑐 ∈ ℝ.
2 {𝑥 ∈ 𝐸 ∣ 𝑓(𝑥) ≥ 𝑐} is measurable for each 𝑐 ∈ ℝ.
3 {𝑥 ∈ 𝐸 ∣ 𝑓(𝑥) < 𝑐} is measurable for each 𝑐 ∈ ℝ.
4 {𝑥 ∈ 𝐸 ∣ 𝑓(𝑥) ≤ 𝑐} is measurable for each 𝑐 ∈ ℝ.

Ryan Azim Shaikh Measure Theory and Integration

Measurable Function

Measurable function
An extended real-valued function 𝑓 with measurable domain is said
to be Lebesgue measurable or simply measurable if for each 𝑐 ∈ ℝ,
the set {𝑥 ∈ 𝐸 ∣ 𝑓(𝑥) > 𝑐} is measurable.

Equivalently the inverse image of every open set in the codomain

should be a measurable set in the domain.

Ryan Azim Shaikh Measure Theory and Integration

Propositions

Proposition
A real-valued function that is continuous on its measurable domain
is measurable.

Proposition
A monotone function that is defined on an interval is measurable.

Ryan Azim Shaikh Measure Theory and Integration

Proposition

The pointwise limit of a sequence of measurable functions is

measurable.

Ryan Azim Shaikh Measure Theory and Integration

Definition

A real-valued function 𝜑 defined on a measurable set 𝐸 is called

simple if it is measurable and takes only a finite number of values.

Ryan Azim Shaikh Measure Theory and Integration

The Simple Approximation lemma

Lemma
Let 𝑓 be a measurable real-valued function on 𝐸. Assume 𝑓 is
bounded on 𝐸, that is, there is an 𝑀 ≥ 0 for which |𝑓| ≤ 𝑀 on 𝐸.
Then for each 𝜀 > 0, there are simple functions 𝜑𝜀 and 𝜓𝜀 defined
on 𝐸 which have the following properties

𝜑𝜀 ≤ 𝑓 ≤ 𝜓𝜀 and 0 ≤ 𝜓𝜀 − 𝜑𝜀 < 𝜀 on 𝐸.

Ryan Azim Shaikh Measure Theory and Integration

The Simple approximation Theorem

Theorem
An extended real-valued function 𝑓 on a measurable set 𝐸 is
measurable iff there is a sequence {𝜑𝑛 }𝑛∈ℕ of simple functions on 𝐸
which converges pointwise on 𝐸 to 𝑓 and has the property that:

|𝜑𝑛 | ≤ |𝑓| on 𝐸 ∀𝑛 ∈ ℕ

if 𝑓 is nonnegative, we may choose {𝜑𝑛 }𝑛∈ℕ to be increasing.

Ryan Azim Shaikh Measure Theory and Integration

Littlewood’s Three Principles

Every measurable set is nearly a finite union of intervals.

Every measurable function is nearly continuous.
Every pointwise convergent sequence of measurable functions
is nearly uniformly convergent.

Ryan Azim Shaikh Measure Theory and Integration

Egoroff’s Theorem

Theorem
Assume 𝐸 has a finite measure. Let {𝑓𝑛 }𝑛∈ℕ be a sequence of
measurable functions on 𝐸 that converges pointwise on 𝐸 to the
real-valued function 𝑓. Then for each 𝜀 > 0, there is a closed set 𝐹
contained in 𝐸, such that

𝑓𝑛 → 𝑓 uniformly on 𝐹 and 𝑚(𝐸 ∼ 𝐹) < 𝜀

Ryan Azim Shaikh Measure Theory and Integration

Lusin’s Theorem

Proposition
Let 𝑓 be a simple function defined on 𝐸. Then for each 𝜀 > 0, there
is a continuous function 𝑔 on ℝ and a closed 𝐹 contained in 𝐸 for
which
𝑓 = 𝑔 on 𝐹 and 𝑚(𝐸 ∼ 𝐹) < 𝜀

Theorem
Let 𝑓 be a real-valued measurable function on 𝐸.Then for each
𝜀 > 0, there is a continuous function 𝑔 on ℝ and a closed set 𝐹
contained in 𝐸 for which

𝑓 = 𝑔 on 𝐹 and 𝑚(𝐸 ∼ 𝐹) < 𝜀

Ryan Azim Shaikh Measure Theory and Integration

Lebesque Measurable Functions

Ryan Azim Shaikh Measure Theory and Integration

The Riemann Integral

Definition
The lower and upper Darboux sums for 𝑓 with respect to 𝑃,
respectively, by
𝑛
𝐿(𝑓, 𝑃) = ∑ 𝑚𝑖 ⋅ (𝑥𝑖 − 𝑥𝑖−1 )
𝑖=1

and 𝑛
𝑈(𝑓, 𝑃) = ∑ 𝑀𝑖 ⋅ (𝑥𝑖 − 𝑥𝑖−1 ).
𝑖=1

Ryan Azim Shaikh Measure Theory and Integration

The Riemann Integral

Definition
The lower and upper Riemann integrals of 𝑓 over [𝑎, 𝑏] are defined,
respectively, as
𝑏
⨜ 𝑓 = sup{𝐿(𝑓, 𝑃)|𝑃 is a partition of [a,b]}.
𝑎

and
𝑏
⨛ 𝑓 = inf{𝑈(𝑓, 𝑃)|𝑃 is a partition of [a,b]}.
𝑎

Ryan Azim Shaikh Measure Theory and Integration

The Riemann Integral

Definition
A bounded real-valued function 𝑓 defined on [𝑎, 𝑏] is said to be
Reimann integrable if it’s upper and lower reimann integral
coincide, further this common value is it’s Reimann integral.
𝑏 𝑏 𝑏
∫ 𝑓=⨜ 𝑓=⨛ 𝑓
𝑎 𝑎 𝑎

Ryan Azim Shaikh Measure Theory and Integration

The Lebesque Integral

Lebesque Integral of a simple function

For a simple function 𝜓 defined on a set 𝐸 of finite measure, we
define the integral of 𝜓 over 𝐸 as
𝑛
∫ 𝜓 = ∑ 𝑎𝑖 ⋅ 𝑚(𝐸𝑖 )
𝐸 𝑖=1

where 𝜓 has the canonical form

𝑛
𝜓 = ∑ 𝑎𝑖 ⋅ 𝜒𝐸𝑖 on 𝐸, where 𝐸𝑖 = {𝑥 ∈ 𝐸|𝜓(𝑥) = 𝑎𝑖 }.
𝑖=1

The Lebesque Integral is:

Linear
Monotonous
Finitely additive over domains

Ryan Azim Shaikh Measure Theory and Integration

The Lebesque Integral

Definition
For a bounded real-valued function defined on a set of finite
measure 𝐸, the lower and upper Lebesque integral are defined,
respectively, as

sup {∫ 𝜓||𝜓 simple and 𝜓 ≤ 𝑓 on 𝐸}

𝐸

and
sup {∫ 𝜓||𝜓 simple and 𝜓 ≤ 𝑓 on 𝐸}
𝐸

𝑓 is said to be Lebesque integrable over 𝐸 provided its upper and

lower lebesque integral coincide. The common value of these
integrals is known as the Lebesque integral of 𝑓 over 𝐸.

Ryan Azim Shaikh Measure Theory and Integration

Properties of the Lebesque Integral
Let 𝑓 be a bounded function defined on the closed, bounded interval
[𝑎, 𝑏]. if 𝑓 is Reimann integrable over [𝑎, 𝑏], then it is lebesque
integrable over [𝑎, 𝑏] and the two integrals are equal.
Let 𝑓 be a bounded measurable function on a set of finite measure
𝐸. Then 𝑓 is integrable over 𝐸.
(Linearity) Let 𝑓 and 𝑔 be bounded measurable functions over a set
𝐸 of finite measure. Then for any real 𝛼 and 𝛽,

∫(𝛼𝑓 + 𝛽𝑔) = 𝛼 ∫ 𝑓 + 𝛽 ∫ 𝑔.
𝐸 𝐸 𝐸

(Monotonicity) Moreover, if 𝑓 ≤ 𝑔 on 𝐸, then

∫ 𝑓 ≤ ∫ 𝑔.
𝐸 𝐸

Let 𝑓 be a bounded measurable function on a set of finite measure

𝐸. If 𝐴 and 𝐵 are disjoint measurable subsets of 𝐸. Then

∫ 𝑓 = ∫ 𝑓 + ∫ 𝑓.
𝐴∪𝐵 𝐴 𝐵

Ryan Azim Shaikh Measure Theory and Integration

The Bounded Convergence Theorem

Theorem
Let {𝑓𝑛 }𝑛∈ℕ be a uniformly bounded sequence of measurable
functions defined on 𝐸, a finite measure set, that converges
pointwise to 𝑓. That is, there exists 𝑀 > 0 such that |𝑓𝑛 | < 𝑀 on 𝐸
for all 𝑛 ∈ ℕ and lim𝑛→∞ 𝑓𝑛 (𝑥) = 𝑓(𝑥) for all 𝑥 in 𝐸, then

lim ∫ 𝑓𝑛 = ∫ 𝑓.
𝑛→∞
𝐸 𝐸

Ryan Azim Shaikh Measure Theory and Integration

The Lebesque Integral

Definition
For a nonnegative measurable function 𝑓 defined on 𝐸, we define
the integral of 𝑓 over 𝐸 by

∫ 𝑓 = sup {∫ 𝑔||𝑔 bounded, measurable, of finite support and 0 ≤ 𝑔 ≤ 𝑓 o

𝐸 𝐸

Ryan Azim Shaikh Measure Theory and Integration

Chebyshev’s Inequality

Theorem
Let 𝑓 be a nonnegative measurable function on 𝐸. Then for any
𝜆 > 0,
1
𝑚{𝑥 ∈ 𝐸|𝑓(𝑥) ≥ 𝜆} ≤ ⋅ ∫ 𝑓.
𝜆 𝐸

Ryan Azim Shaikh Measure Theory and Integration

Properties of the Lebesque Integral

(Linearity) Let 𝑓 and 𝑔 be nonnegative measurable functions over a

set 𝐸. Then for any real 𝛼 and 𝛽,

∫(𝛼𝑓 + 𝛽𝑔) = 𝛼 ∫ 𝑓 + 𝛽 ∫ 𝑔.
𝐸 𝐸 𝐸

(Monotonicity) Moreover, if 𝑓 ≤ 𝑔 on 𝐸, then

∫ 𝑓 ≤ ∫ 𝑔.
𝐸 𝐸

Let 𝑓 be a nonnegative measurable function on a set 𝐸. If 𝐴 and 𝐵

are disjoint measurable subsets of 𝐸. Then

∫ 𝑓 = ∫ 𝑓 + ∫ 𝑓.
𝐴∪𝐵 𝐴 𝐵

Ryan Azim Shaikh Measure Theory and Integration

Fatou’s Lemma

Lemma
Let {𝑓𝑛 }𝑛∈ℕ be a sequence of nonnegative measurable functions on
𝐸. If 𝑓𝑛 → 𝑓 pointwise a.e. on 𝐸, then

∫ 𝑓 ≤ lim inf ∫ 𝑓𝑛 .
𝑛→∞
𝐸 𝐸

Ryan Azim Shaikh Measure Theory and Integration

Monotone Convergence Theorem

Theorem
Let {𝑓𝑛 }𝑛∈ℕ be an increasing sequence of nonnegative measurable
functions on 𝐸. If 𝑓𝑛 → 𝑓 pointwise a.e. on 𝐸, then

lim ∫ 𝑓𝑛 = ∫ 𝑓.
𝑛→∞
𝐸 𝐸

Ryan Azim Shaikh Measure Theory and Integration

The General Lebesque Integral
Definition
For an extended real-valued function 𝑓 on 𝐸, we define the positive
(𝑓+ ) and negative (𝑓− ) part of 𝑓, respectively as

𝑓+ (𝑥) = max{𝑓(𝑥), 0} and 𝑓− (𝑥) = 𝑚𝑎𝑥{−𝑓(𝑥), 0} ∀𝑥 ∈ 𝐸.

Proposition
Let 𝑓 be a measurable function on 𝐸. Then 𝑓+ and 𝑓− are
integrable over 𝐸 iff |𝑓| is integrable over 𝐸.

The General Lebesque Integral

A measurable function 𝑓 is said to be integrable over 𝐸 if |𝑓| is
integrable over 𝐸. We define the integral of 𝑓 over 𝐸 as

∫ 𝑓 = ∫ 𝑓+ − ∫ 𝑓− .
𝐸 𝐸 𝐸

Ryan Azim Shaikh Measure Theory and Integration