Study guide: Product Integration and Change of Variables

Course Outline

  1. Product Sigma-Algebras and Sections
  2. Measurable Sections and Sigma-Finiteness
  3. Product measure
  4. Independence of Random Variables
  5. Integration Against Product Measures
  6. Lebesgue Measure in Dimension d
  7. Change of Variables Formula
  8. Fubini Theorems
  9. Independence and Expectations
  10. The Characteristic Function
  11. Independence Through CDFs and Densities

1. Product Sigma-Algebras and Sections

Key Concepts & Definitions

  • Product Οƒ-algebra: Οƒ-algebra generated by the measurable rectangles A Γ— B with A ∈ E₁ and B ∈ Eβ‚‚.
  • x-section: For C βŠ† E₁ Γ— Eβ‚‚ and x ∈ E₁, the x-section of C is Cβ‚“ = {y ∈ Eβ‚‚ : (x, y) ∈ C}.

β˜… Must-know

  • If C belongs to E₁ βŠ— Eβ‚‚, then Cβ‚“ belongs to Eβ‚‚ for every x ∈ E₁, and CΚΈ belongs to E₁ for every y ∈ Eβ‚‚.

Further detail

  • For a finite collection of measurable spaces (Eα΅’, Eα΅’), the product Οƒ-algebra on E₁ Γ— β‹― Γ— Eβ‚™ is generated by the rectangles A₁ Γ— β‹― Γ— Aβ‚™ with Aα΅’ ∈ Eα΅’.

2. Measurable Sections and Sigma-Finiteness

Key Concepts & Definitions

  • Οƒ-finite measure: a measure on (E, E) for which there is an increasing sequence of measurable sets (Xβ‚™) such that ⋃ₙ Xβ‚™ = E and ΞΌ(Xβ‚™) < ∞ for every n.
  • Οƒ-finite measure: A measure ΞΌ on a measurable space (E, 𝓔) is Οƒ-finite if there are measurable sets Xβ‚™ increasing to E and satisfying ΞΌ(Xβ‚™) < ∞ for every n.

Essential Points

  • If f : (E₁ Γ— Eβ‚‚, E₁ βŠ— Eβ‚‚) β†’ (G, G) is measurable, then for every x ∈ E₁ the function fβ‚“(y) = f(x, y) is Eβ‚‚-measurable, and for every y ∈ Eβ‚‚ the function fΚΈ(x) = f(x, y) is E₁-measurable.

3. Product measure

Key Concepts & Definitions

  • Product Measure: The measure mm on E1Γ—E2E_1\times E_2 is defined by m(C)=∫E1Ο‰(Cx) μ(dx)m(C)=\int_{E_1}\omega(C_x)\,\mu(dx) for measurable sets CC.

β˜… Must-know

  • If ΞΌ and Ξ½ are Οƒ-finite measures on (E₁, 𝓔₁) and (Eβ‚‚, 𝓔₂), respectively, there exists a unique measure ΞΌ βŠ— Ξ½ on (E₁ Γ— Eβ‚‚, 𝓔₁ βŠ— 𝓔₂) such that ΞΌ βŠ— Ξ½(A Γ— B) = ΞΌ(A)Ξ½(B) for every A ∈ 𝓔₁ and B ∈ 𝓔₂, with the convention 0 Γ— ∞ = 0.

πŸ“ Formula - For every C ∈ 𝓔₁ βŠ— 𝓔₂, ΞΌβŠ—Ξ½(C)=∫E1Ξ½(Cx) μ(dx)=∫E2ΞΌ(Cy) ν(dy).\mu \otimes \nu(C) = \int_{E_1} \nu(C_x)\,\mu(dx) = \int_{E_2} \mu(C_y)\,\nu(dy).

πŸ“ Formula - For measurable rectangles, the product measure satisfies m(AΓ—B)=ΞΌ(A)Ο‰(B)m(A\times B)=\mu(A)\omega(B) for A∈E1A\in E_1 and B∈E2B\in E_2.

Further detail

  • The product measure ΞΌ βŠ— Ξ½ of two Οƒ-finite measures is Οƒ-finite.

  • For every C ∈ 𝓔₁ βŠ— 𝓔₂, the section measure x ↦ Ξ½(Cβ‚“) is 𝓔₁-measurable.

  • The construction of mm is completed by proving that it is a measure; this proof is left as an exercise.

4. Independence of Random Variables

Key Concepts & Definitions

  • Independence of Random Variables: Random variables X1,…,XnX_1,\ldots,X_n are independent if, for every choice of measurable sets Bk∈EkB_k\in E_k and every subset JβŠ†{1,…,n}J\subseteq\{1,\ldots,n\}, P(β‹‚k∈J{Xk∈Bk})=∏k∈JP(Xk∈Bk)P(\bigcap_{k\in J}\{X_k\in B_k\})=\prod_{k\in J}P(X_k\in B_k).
  • Independence of Events: Events A1,…,AnA_1,\ldots,A_n are independent if, for every subset JβŠ†{1,…,n}J\subseteq\{1,\ldots,n\}, P(β‹‚k∈JAk)=∏k∈JP(Ak)P(\bigcap_{k\in J}A_k)=\prod_{k\in J}P(A_k).
  • Joint Law: The law of (X,Y)(X,Y) is the probability measure defined on measurable sets CC by P(X,Y)(C)=P((X,Y)∈C)P_{(X,Y)}(C)=P((X,Y)\in C).

β˜… Must-know

πŸ“Œ Two random variables XX and YY are independent if and only if their joint law satisfies P(X,Y)=PXβŠ—PYP_{(X,Y)}=P_X\otimes P_Y.

Further detail

  • A pair-valued map (X,Y)(X,Y) is measurable into the product measurable space if and only if XX and YY are measurable into their respective spaces.

  • For finitely many random variables, independence is equivalent to the joint law being the product of their individual laws, and the product of measures is associative.

Memory Hook

Event independence concerns events; random-variable independence concerns their inverse-image events.

5. Integration Against Product Measures

β˜… Must-know

πŸ“ Formula - For a nonnegative simple function f=βˆ‘k=1nΞ±k1Akf=\sum_{k=1}^{n}\alpha_k\mathbf{1}_{A_k} on a product space, its integral against ΞΌ\mu is ∫f dΞΌ=βˆ‘k=1nΞ±kΞΌ(Ak)\int f\,d\mu=\sum_{k=1}^{n}\alpha_k\mu(A_k), where the sets AkA_k are measurable and the coefficients Ξ±k\alpha_k are nonnegative.

πŸ“ Formula - For a nonnegative measurable function ff, its integral against ΞΌ\mu is ∫f dΞΌ=sup⁑{∫g dΞΌ:gΒ isΒ simpleΒ andΒ g≀f}\int f\,d\mu=\sup\{\int g\,d\mu:g\text{ is simple and }g\leq f\}.

πŸ“Œ A measurable function ff belongs to L1L^1 for the measure ΞΌ\mu exactly when ∫∣fβˆ£β€‰dΞΌ<+∞\int |f|\,d\mu<+\infty.

Further detail

πŸ“ Formula - For an integrable measurable function ff, its integral is defined by ∫f dΞΌ=∫f+ dΞΌβˆ’βˆ«fβˆ’β€‰dΞΌ\int f\,d\mu=\int f^+\,d\mu-\int f^-\,d\mu, where f+=max⁑(f,0)f^+=\max(f,0) and fβˆ’=βˆ’min⁑(f,0)f^-=-\min(f,0).

Memory Hook

Simple functions first, nonnegative limits next, signed integrable functions last.

6. Lebesgue Measure in Dimension d

Key Concepts & Definitions

  • Volume of a d-cell: The volume of the d-cell W=∏k=1d]ak,bk[W=\prod_{k=1}^{d}]a_k,b_k[ is Vol⁑(W)=∏k=1d(bkβˆ’ak)\operatorname{Vol}(W)=\prod_{k=1}^{d}(b_k-a_k) when ak<bka_k<b_k for every coordinate.

β˜… Must-know

  • The product measure Ξ»βŠ—d\lambda^{\otimes d} is the unique measure satisfying Ξ»βŠ—d(A1Γ—β‹―Γ—Ad)=∏k=1dΞ»(Ak)\lambda^{\otimes d}(A_1\times\cdots\times A_d)=\prod_{k=1}^{d}\lambda(A_k) for Borel sets A1,…,AdA_1,\ldots,A_d.

  • The Lebesgue measure on Rd\mathbb{R}^d is the restriction of the outer measure to its measurable sigma-algebra, which contains the Borel sigma-algebra and is its completion with respect to Lebesgue measure.

πŸ“Œ Lebesgue measure on Rd\mathbb{R}^d is the unique measure on the Borel sigma-algebra that assigns each d-cell its volume.

Further detail

  • The Borel sigma-algebra on Rd\mathbb{R}^d is the product of dd copies of the Borel sigma-algebra on R\mathbb{R}.

πŸ“ Formula - For any subset AβŠ†RdA\subseteq\mathbb{R}^d, the outer measure is Ξ»(d,βˆ—)(A)=inf⁑{βˆ‘jVol⁑(Wj):AβŠ†β‹ƒjWj,Β WjΒ areΒ d-cells}\lambda^{(d,*)}(A)=\inf\left\{\sum_j\operatorname{Vol}(W_j):A\subseteq\bigcup_j W_j,\ W_j\text{ are d-cells}\right\}.

  • The d-fold product of one-dimensional Lebesgue measure equals Lebesgue measure on Rd\mathbb{R}^d.

7. Change of Variables Formula

Key Concepts & Definitions

  • C1-diffeomorphism: A map Ο†:Vβ†’U\varphi:V\to U between open subsets of Rd\mathbb{R}^d is a C1-diffeomorphism if it is bijective and both it and its inverse are differentiable with continuous differentials.
  • Jacobian determinant: The Jacobian determinant of a differentiable map Ο†:Vβ†’Rd\varphi:V\to\mathbb{R}^d is JΟ†(x)=det⁑(βˆ‚Ο†jβˆ‚xi(x))i,jJ_\varphi(x)=\det\left(\frac{\partial\varphi_j}{\partial x_i}(x)\right)_{i,j}.

β˜… Must-know

πŸ“Œ For a C1-diffeomorphism Ο†:Vβ†’U\varphi:V\to U and measurable f:Uβ†’Rf:U\to\mathbb{R}, ff is integrable on UU if and only if ∣JΟ†βˆ£(fβˆ˜Ο†)|J_\varphi|(f\circ\varphi) is integrable on VV.

πŸ“ Formula - For a CΒΉ-diffeomorphism Ο–:Vβ†’U and an integrable function f on U, ∫Uf(y) ϑ(dy)=∫V∣JΟ–(x)∣f(Ο–(x)) ϑ(dx)\int_U f(y)\,\vartheta(dy)=\int_V |JΟ–(x)|f(Ο–(x))\,\vartheta(dx), and the equality also holds for measurable nonnegative f with values in [0,+∞].

πŸ“Œ For a CΒΉ-diffeomorphism Ο–:Vβ†’U, f is integrable on U if and only if |JΟ–|(fβˆ˜Ο–) is integrable on V.

Further detail

πŸ“Œ The change-of-variables integral equality also holds with values in [0,+∞][0,+\infty] for every measurable nonnegative function on UU.

  • The change-of-variables proof extends the result from indicator functions to simple functions, then to nonnegative measurable functions, and finally to integrable functions.

  • The admitted first step of the change-of-variables proof states that for any cuboid CβŠ‚V, Ο‚(Ο–(C))=∫_C |JΟ–| dΟ‚.

8. Fubini Theorems

β˜… Must-know

πŸ“Œ For Οƒ-finite measures ΞΌ and Ξ΅ and a measurable function f:E₁×Eβ‚‚β†’[0,+∞], the functions xβ†¦βˆ«{Eβ‚‚}f(x,y)Ξ΅(dy) and yβ†¦βˆ«{E₁}f(x,y)ΞΌ(dx) are measurable.

πŸ“ Formula - For Οƒ-finite measures ΞΌ and Ξ΅ and measurable f:E₁×Eβ‚‚β†’[0,+∞], ∫E1Γ—E2f d(ΞΌβŠ—Ξ΅)=∫E1(∫E2f(x,y) Ρ(dy))ΞΌ(dx)=∫E2(∫E1f(x,y) μ(dx))Ξ΅(dy).\int_{E₁\times Eβ‚‚}f\,d(\mu{}\otimes Ξ΅)=\int_{E₁}\left(\int_{Eβ‚‚}f(x,y)\,Ξ΅(dy)\right)\mu{}(dx)=\int_{Eβ‚‚}\left(\int_{E₁}f(x,y)\,\mu{}(dx)\right)Ξ΅(dy).

πŸ“Œ For f integrable with respect to ΞΌβŠ—Ξ΅, its sections y↦f(x,y) are integrable for ΞΌ-almost every x, and its sections x↦f(x,y) are integrable for Ξ΅-almost every y.

Further detail

  • For integrable f with respect to ΞΌβŠ—Ξ΅, the two iterated integrals are well-defined except possibly on null sets and define integrable functions on their respective spaces.

πŸ“ Formula - For f integrable with respect to ΞΌβŠ—Ξ΅, ∫E1Γ—E2f d(ΞΌβŠ—Ξ΅)=∫E1(∫E2f(x,y) Ρ(dy))ΞΌ(dx)=∫E2(∫E1f(x,y) μ(dx))Ξ΅(dy).\int_{E₁\times Eβ‚‚}f\,d(\mu{}\otimes Ξ΅)=\int_{E₁}\left(\int_{Eβ‚‚}f(x,y)\,Ξ΅(dy)\right)\mu{}(dx)=\int_{Eβ‚‚}\left(\int_{E₁}f(x,y)\,\mu{}(dx)\right)Ξ΅(dy).

Memory Hook

Tonelli covers nonnegative functions; Lebesgue covers integrable functions.

9. Independence and Expectations

β˜… Must-know

  • Random variables X and Y are independent if and only if E[f(X)g(Y)]=E[f(X)]E[g(Y)] for every pair of nonnegative measurable functions f and g, or equivalently for every pair of bounded measurable functions f and g.

  • For measurable non-negative bounded functions f and g, independence of X and Y implies E[f(X)g(Y)] = E[f(X)]E[g(Y)].

  • If E[f(X)g(Y)] = E[f(X)]E[g(Y)] for all measurable non-negative bounded f and g, then X and Y are independent, as shown by taking f and g to be indicator functions of measurable sets.

Further detail

πŸ“ Formula - For independent Eβ‚–-valued random variables X₁,…,Xβ‚™ and nonnegative measurable functions fβ‚–, E[∏k=1nfk(Xk)]=∏k=1nE[fk(Xk)]\mathbb{E}\left[\prod_{k=1}^{n}f_k(X_k)\right]=\prod_{k=1}^{n}\mathbb{E}[f_k(X_k)]; the identity also holds when all fβ‚– are bounded and measurable.

  • For independent random variables X and Y, the factorization of expectations also holds for bounded measurable complex-valued functions f and g.

  • The equivalence between independence and factorization of expectations extends to n random variables by applying Fubini nβˆ’1 times.

10. The Characteristic Function

Key Concepts & Definitions

  • Characteristic function: The characteristic function of a random vector X in β„α΅ˆ is Ο•_X(t) = E[e^{i⟨t,X⟩}] = ∫_{β„α΅ˆ} e^{i⟨t,x⟩} P_X(dx), for t ∈ β„α΅ˆ.

β˜… Must-know

πŸ“Œ The components X₁, …, X_d of a random vector are independent if and only if Ο•_X(t) = ∏{k=1}^d Ο•{X_k}(t_k) for every t = (t₁, …, t_d) ∈ β„α΅ˆ.

Further detail

  • For every t ∈ β„α΅ˆ, the characteristic function satisfies |Ο•_X(t)| ≀ 1 because |e^{i⟨t,x⟩}| = 1.

11. Independence Through CDFs and Densities

Key Concepts & Definitions

  • Joint cumulative distribution function: The joint cumulative distribution function of real-valued random variables X and Y is F_{(X,Y)}(x,y) = P(X ≀ x, Y ≀ y), for (x,y) ∈ ℝ².

β˜… Must-know

πŸ“Œ Real-valued random variables X and Y are independent if and only if F_{(X,Y)}(x,y) = F_X(x)F_Y(y) for every (x,y) ∈ ℝ².

πŸ“Œ When the joint law has density f_{(X,Y)}, X and Y are independent if and only if f_{(X,Y)}(x,y) = f_X(x)f_Y(y) for Lebesgue-almost every (x,y) ∈ ℝ².

Further detail

  • If the joint law of (X,Y) has density f_{(X,Y)} with respect to Lebesgue measure on ℝ², then its marginal densities are f_X(x) = ∫ f_{(X,Y)}(x,y) dy and f_Y(y) = ∫ f_{(X,Y)}(x,y) dx, and these integrals are finite outside a Lebesgue-null set.

Memory Hook

CDF factorization tests independence directly; density factorization does so almost everywhere.

Test yourself

Take a 30-question practice quiz on Product Integration and Change of Variables, with every answer explained.

1. How is the product Οƒ-algebra on E1Γ—E2E_1 \times E_2 defined?

2. Suppose C∈E1βŠ—E2C \in \mathcal{E}_1 \otimes \mathcal{E}_2. What can be concluded about its sections for each fixed x∈E1x \in E_1 and y∈E2y \in E_2?

Take the quiz β†’

Study with flashcards

Memorize the key concepts of Product Integration and Change of Variables with 65 flashcards.

What generates the product Οƒ-algebra E1βŠ—E2E_1 \otimes E_2 on E1Γ—E2E_1 \times E_2?

Measurable rectangles AΓ—BA \times B with A∈E1A \in E_1 and B∈E2B \in E_2.

If C∈E1βŠ—E2C \in E_1 \otimes E_2, where does each x-section CxC_x belong?

E2E_2.

If C∈E1βŠ—E2C \in E_1 \otimes E_2, where does each y-section CyC^y belong?

E1E_1.

See the flashcards β†’

Related classes

Make your own study guides

Upload your class notes and Revizly writes a study guide, a quiz and flashcards in about a minute.

Study guide maker