M anam 0 an e c unan e c the old postulates are

Info icon This preview shows pages 50–53. Sign up to view the full content.

m, AnAm = 0, An E .C UnAn E .C. The old postulates are equivalent to the new ones. Here we only check that old implies new. Suppose >..;, >..;, >..) are true. Then .l..t is true. Since Q E .C, if A E .C, then A C Q and by>..;, Q \A = Ac E .C, which shows that >..2 is true. If A, B E .C are disjoint, we show that A U B E .C. Now Q \ A E .C and B C Q \ A (since (J) E B implies (J) fl. A which means (J) E Ac = Q \A) so by>..; we have (Q \A)\ B = AcBc E .C and by >..2 we have (AcBc)c =AU BE .C which is AJ for finitely many sets. Now if A j E .Care mutually disjoint for j = 1, 2, ... ,
Image of page 50

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

2.2 More on Closure 37 define Bn = U}=tAi . Then Bn e £by the prior argument for 2 sets and by >..3 we have UnBn = limn-+oo t Bn E £. Since UnBn = UnAn we have UnAn E £ which is AJ. 0 Remark. It is clear that a a-field is always a >..-system since the new postulates obviously hold. Recall that a rr - system is a class of sets closed under finite intersections; that is, 'Pis a rr-system if whenever A, B e 'P we have AB e 'P . We are now in a position to state Dynkin's theorem. Theorem 2.2.2 (Dynkin's theorem) (a) lf'P is a rr-system and£ is a >..-system such that P c £, then a ('P) C £. (b) lf'P is a rr-system a('P) = £('P), that is, the minimal a-field over P equals the minimal >.. - system over P. Note (b) follows from (a). To see this assume (a) is true. Since 'P c C('P), we have from (a) that a('P) c C('P). On the other hand, a('P), being a a-field, is a >..-system containing P and hence contains the minimal >..-system over 'P, so that a ('P) ::> £('P). Before the proof of (a), here is a significant application of Dynkin's theorem. Proposition 2.2.3 Let Pt. Pz be two probability measures on (Q, 8) . The class £ : ={A e 8: Pt(A) = Pz(A)} is a >..-system. Proof of Proposition 2.2.3. We show the new postulates hold : (J..t) n E £since Pt(fl) = Pz(n) = 1. (J..z) A e £implies Ace£, since A E £means Pt(A) = Pz(A), from which (J..3) If {A i} is a mutually disjoint sequence of events in £, then P 1 (A i) = Pz (A i) for all j, and hence P 1 (UAj) = LPt(Aj) = L:Pz(Aj) = Pz<UAj) j j j j so that 0
Image of page 51
38 2. Probability Spaces Corollary 2.2.1 If Pt. P2 are two probability measures on (Q, B) and ifP is a ;r-system such that 'v'A e P: Pt(A) = P2(A), then 'v'B e a(P): Pt(B) = P2(B). Proof of Corollary 2.2.1. We have C ={A e B: Pt(A) = P2(A)} is a A-system. But C :::> P and hence by Dynkin's theorem C :::> a(P) . 0 Corollary 2.2.2 Let n = JR. Let Pt, P2 be two probability measures on (JR., B(IR.)) such that their distribution functions are equal: 'v'x E JR.: Ft(X) = Pt((-oo,x)) = F2(x) = P2((-oo,x]) . Then on B(IR.). So a probability measure on JR. is uniquely determined by its distribution func- tion. Proof of Corollary 2.2.2. Let P = {(-oo,x]: x e JR.}. Then P is a ;r -system since (-oo,x) n (-oo,y] = (-oo,x Ay) E P. Furthermore a (P) = B(IR.) since the Borel sets can be generated by the semi- infinite intervals (see Section 1.7). SoFt (x) = F2(x) for all x e JR., means Pt = P2 on P and hence Pt = P2 on a(P) = B(IR.). 0 2.2.2 Proof of Dynkin 's theorem Recall that we only need to prove: If Pis a ;r-system and Cis a A-system then P C C implies a (P) C C.
Image of page 52

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

Image of page 53
This is the end of the preview. Sign up to access the rest of the document.

{[ snackBarMessage ]}

What students are saying

  • Left Quote Icon

    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.

    Student Picture

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

  • Left Quote Icon

    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.

    Student Picture

    Dana University of Pennsylvania ‘17, Course Hero Intern

  • Left Quote Icon

    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.

    Student Picture

    Jill Tulane University ‘16, Course Hero Intern