A Probability Path.pdf

# E 2 n 0 53 limits and integrals this section presents

• Test Prep
• 464

This preview shows pages 145–149. Sign up to view the full content.

E 2 n 0 5.3 Limits and Integrals This section presents a sequence of results which describe how expectation and limits interact. Under certain circumstances we are allowed to interchange expec- tation and limits. In this section we will learn when this is safe. Theorem 5.3.1 (Monotone Convergence Theorem (MCT)) If 0 Xn t X then 0 E(Xn) t E(X). This was proved in the previous subsection 5.2.3. See 3 page 123. Corollary 5.3.1 (Series Version ofMCT) If 0 are non-negative random variables for n 1, then 00 00 = j=l j=l so that the expectation and infinite sum can be interchanged. To see this, just write n = lim t £(" n_.oo j=l (MCT) n = lim t n_.oo j=l 00 = j=l 0

This preview has intentionally blurred sections. Sign up to view the full version.

132 5. Integration and Expectation Theorem 5.3.2 (Fatou Lemma) If X n 0, then £(lim infXn) ::::: lim inf E(Xn). n-+oo n-+00 More generally, if there exists Z e Lt and Xn Z, then E(liminfXn)::::: liminfE(Xn) . n-+oo n-+oo Proof of Fatou. If X n 0, then E (lim infXn) = E t (6 x,)) n-+00 = t E (3xk) (from MCI' 5.3.1) ::::: lim inf E(Xn). n-+oo For the case where we assume X n Z, we have X n - Z 0 and E (liminf(Xn- Z))::::: liminfE(Xn- Z) n-+00 n-+oo so E(liminfXn)- E(Z)::::: liminfE(Xn)- E(Z). n-+oo n-+oo The result follows by cancelling E(Z) from both sides of the last relation. 0 Corollary 5.3.2 (More Fatou) If Xn ::::: Z where Z e Lt, then E(limsupXn) limsupE(Xn). n-+oo n-+oo Proof. This follows quickly from the previous Fatou Lemma 5.3.2. If Xn ::::: Z, then -Xn -Z e L., and the Fatou Lemma 5.3.2 gives E(liminf(-Xn))::::: liminf£(-Xn), n-+00 n-+oo so that £(-liminf(-Xn)) -liminf(-EXn) . n-+oo n-+oo The proof is completed by using the relation - lim inf- = lim sup . 0 Canonical Example. This example is typical of what can go wrong when limits and integrals are interchanged without any dominating condition. Usually some- thing very nasty happens on a small set and the degree of nastiness overpowers the degree of smallness.
5.3 Limits and Integrals 133 Let (Q, B, P) = ([0, 1], 8((0, 1]), .l.) where, as usual, ). is Lebesgue measure . Define Xn = n 2 1(0.1/n)· For any we (0 , 1] , so However so and 1(0, 1/n)(w) --+ 0, Xn--+ 0. 1 E(Xn) = n 2 · - = n--+ oo, n E(liminfXn) = 0 < liminf(EXn) = oo n-+00 n-+oo E(limsupXn) = 0, limsupE(Xn) = 00. n-+00 n-+oo So the second part of the Fatou Lemma given in Corollary 5.3 .2 fails. So obviously we cannot hope for Corollary 5. 3.2 to hold without any restriction. 0 Theorem 5.3.3 (Dominated Convergence Theorem (DCT)) If Xn--+ X, and there exists a dominating random variable Z E L 1 such that IXnl then E(Xn)--+ E(X) and EIXn- XI--+ 0. Proof of DCT. This is an easy consequence of the Fatou Lemma. We have -Z ::5 Xn ::5 Z and - Z e L 1 as well as Z e L So both parts of Fatou 's lemma apply: E(X) =E(liminfXn) n-+oo :::;lim inf E(Xn) n-+oo :::;limsupE(Xn) n-+oo supXn) n-+oo =E(X). (Fatou Lemma 5.3.2) (since inf < sup) (Corollary 5.3. 2) Thus all inequalities are equality. The rest follows from IX n - X I ::::; 2Z . o

This preview has intentionally blurred sections. Sign up to view the full version.

134 5. Integration and Expectation 5.4 Indefinite Integrals Indefinite integrals allow for integration over only part of the Q-space. They are simply defined in terms of indicator functions.
This is the end of the preview. Sign up to access the rest of the document.
• Fall '08
• Staff

{[ snackBarMessage ]}

### What students are saying

• As a current student on this bumpy collegiate pathway, I stumbled upon Course Hero, where I can find study resources for nearly all my courses, get online help from tutors 24/7, and even share my old projects, papers, and lecture notes with other students.

Kiran Temple University Fox School of Business ‘17, Course Hero Intern

• I cannot even describe how much Course Hero helped me this summer. It’s truly become something I can always rely on and help me. In the end, I was not only able to survive summer classes, but I was able to thrive thanks to Course Hero.

Dana University of Pennsylvania ‘17, Course Hero Intern

• The ability to access any university’s resources through Course Hero proved invaluable in my case. I was behind on Tulane coursework and actually used UCLA’s materials to help me move forward and get everything together on time.

Jill Tulane University ‘16, Course Hero Intern