Dynkin system: Difference between revisions

From Wikipedia, the free encyclopedia
Content deleted Content added
SieBot (talk | contribs)
m robot Adding: fi:Dynkinin luokka
Nm420 (talk | contribs)
merged material from "lambda-system", some other minor edits
Line 1: Line 1:
A '''Dynkin system''' on <math>\Omega</math> is a [[set]] <math>\mathcal{D}</math> consisting of certain [[subset]]s of <math>\Omega</math> such that
A '''Dynkin system''', named in honor of the Russian mathematician [[Eugene Dynkin]], is a [[class (set theory)|collection]] of [[subset]]s of another universal [[set]] <math>\Omega</math> satisfying some specific rules. They are also referred to as '''&lambda;-systems'''.


== Definitions ==
*the set <math>\Omega</math> itself is in <math>\mathcal{D}</math>
*if <math>A,B \in \mathcal{D}</math> and <math>A \subseteq B</math> then <math>B \setminus A \in \mathcal{D}</math>
Let <math>\Omega</math> be a [[nonempty]] set, and let <math>D</math> be a collection of subsets of <math>\Omega</math>, i.e. <math>D</math> is a subset of the [[power set]] of <math>\Omega</math>. Then <math>D</math> is a '''Dynkin system''' if
* the set <math>\Omega</math> itself is in <math>D</math>
*if <math>A_n</math> is a [[sequence]] of sets in <math>\mathcal{D}</math> which is increasing in the sense that <math>A_n \subseteq A_{n+1},\ n \ge 1</math>, then the union <math>\bigcup_{k=1}^{\infty}A_k</math> also lies in <math>\mathcal{D}.</math>
* <math>D</math> is [[closure (mathematics)|closed]] under [[relative complement|relative complementation]], i.e. <math>A,B\in D</math> and <math>A \subseteq B</math> implies <math>B \setminus A \in D</math>
* <math>D</math> is closed under the [[countable]] [[union (set theory)|union]] of increasing [[sequence]]s, i.e. <math>A_n\in D</math> and <math>A_n \subseteq A_{n+1},\ n \ge 1</math> implies <math>\cup_{n=1}^{\infty}A_n\in D</math>.


<math>D</math> is a '''&lambda;-system''' if
If <math>\mathcal{J}</math> is any set of subsets of <math>\Omega</math>, then the [[Intersection (set theory)|intersection]] of all the Dynkin systems containing <math>\mathcal{J}</math> is itself a Dynkin system. It is called the Dynkin system generated by <math>\mathcal{J}</math>. It is the smallest Dynkin system containing <math>\mathcal{J}</math>.
* the set <math>\Omega</math> itself is in <math>D</math>
* <math>D</math> is closed under [[complement (set theory)|complementation]], i.e. <math>A\in D</math> implies <math>A^c\in D</math>
* <math>D</math> is closed under [[disjoint sets|disjoint]] countable unions, i.e. <math>A_n\in D, n\geq1</math> with <math>A_i\cap A_j=\emptyset</math> for all <math>i\neq j</math> implies <math>\cup_{n=1}^\infty A_n\in D</math>.


It can be shown that these two definitions are [[logical equivalence|logically equivalent]], so that Dynkin systems are &lambda;-systems and vice versa.
A Dynkin system which is also a [[pi system|&pi;-system]] is a [[sigma algebra|&sigma;-algebra]].


A Dynkin system which is also a [[pi system|&pi;-system]] is a [[sigma algebra|&sigma;-algebra]].
Dynkin systems are named after the Russian mathematician [[Eugene Dynkin]].


Given any collection <math>\mathcal{J}</math> of subsets of <math>\Omega</math>, there exists a unique Dynkin system denoted <math>D\{\mathcal J\}</math> which is minimal with respect to containing <math>\mathcal J</math>. That is, if <math>\tilde D</math> is any Dynkin system containing <math>\mathcal J</math>, then <math>D\{\mathcal J\}\subseteq\tilde D</math>. <math>D\{\mathcal J\}</math> is called the Dynkin system generated by <math>\mathcal{J}</math>. Note <math>D\{\emptyset\}=\{\emptyset,\Omega\}</math>. For another example, let <math>\Omega=\{1,2,3,4\}</math> and <math>\mathcal J=\{1\}</math>; then <math>D\{\mathcal J\}=\{\emptyset,\{1\},\{2,3,4\},\Omega\}</math>.
The '''Dynkin system theorem''' ([[monotone class theorem]], [[Dynkin's lemma]]) states:


== Dynkin's &pi;-&lambda; Theorem ==
Let <math>\mathcal{C}</math> be a &pi;-system; that is, a collection of subsets of <math>\Omega</math> which is closed under pairwise intersections. If <math>\mathcal{D}</math> is a Dynkin system containing <math>\mathcal{C}</math>, then <math>\mathcal{D}</math> also contains the <math>\sigma</math>-algebra <math>\sigma(\mathcal{C})</math> generated by <math>\mathcal{C}</math>.
If <math>P</math> is a &pi;-system and <math>D</math> is a Dynkin system with <math>P\subseteq D</math>, then <math>\sigma\{P\}\subseteq D</math>. In other words, the &\sigma;-algebra generated by <math>P</math> is contained in <math>D</math>.


One application of Dynkin's lemma is the uniqueness of the [[Lebesgue measure]]:
One application of Dynkin's &pi;-&lambda; theorem is the uniqueness of the [[Lebesgue measure]]:


Let (&Omega;, ''B'', &lambda;) be the [[unit interval]] [0,1] with the Lebesgue measure on [[Borel sets]]. Let &mu; be another [[Measure (mathematics)|measure]] on &Omega; satisfying &mu;[(''a'',''b'')] = ''b'' - ''a'', and let ''D'' be the family of sets such that &mu;[D] = &lambda;[D]. Let ''I'' = { (''a'',''b''),[''a'',''b''),(''a'',''b''],[''a'',''b''] : 0 < ''a'' &le; ''b'' < 1 }, and observe that ''I'' is closed under finite intersections, that ''I'' &sub; ''D'', and that ''B'' is the &sigma;-algebra generated by ''I''. One easily shows ''D'' satisfies the above conditions for a Dynkin-system. From Dynkin's lemma it follows that ''D'' is in fact all of ''B'', which is equivalent to showing that the Lebesgue measure is unique.
Let (&Omega;, ''B'', &lambda;) be the [[unit interval]] [0,1] with the Lebesgue measure on [[Borel sets]]. Let &mu; be another [[Measure (mathematics)|measure]] on &Omega; satisfying &mu;[(''a'',''b'')] = ''b'' - ''a'', and let ''D'' be the family of sets such that &mu;[D] = &lambda;[D]. Let ''I'' = { (''a'',''b''),[''a'',''b''),(''a'',''b''],[''a'',''b''] : 0 < ''a'' &le; ''b'' < 1 }, and observe that ''I'' is closed under finite intersections, that ''I'' &sub; ''D'', and that ''B'' is the &sigma;-algebra generated by ''I''. One easily shows ''D'' satisfies the above conditions for a Dynkin-system. From Dynkin's lemma it follows that ''D'' is in fact all of ''B'', which is equivalent to showing that the Lebesgue measure is unique.


== Bibliography ==
{{cite book
| last = Gut
| first = Allan
| title = Probability: A Graduate Course
| publisher = Springer
| date = 2005
| location = New York
| doi = 10.1007/b138932
| isbn = 0-387-22833-0
}}

{{cite book
| last = Billingsley
| first = Patrick
| title = Probability and Measure
| publisher = John Wiley & Sons, Inc.
| date = 1995
| location = New York
| isbn = 0-471-00710-2
}}


----
----

Revision as of 02:41, 4 May 2008

A Dynkin system, named in honor of the Russian mathematician Eugene Dynkin, is a collection of subsets of another universal set satisfying some specific rules. They are also referred to as λ-systems.

Definitions

Let be a nonempty set, and let be a collection of subsets of , i.e. is a subset of the power set of . Then is a Dynkin system if

  • the set itself is in
  • is closed under relative complementation, i.e. and implies
  • is closed under the countable union of increasing sequences, i.e. and implies .

is a λ-system if

  • the set itself is in
  • is closed under complementation, i.e. implies
  • is closed under disjoint countable unions, i.e. with for all implies .

It can be shown that these two definitions are logically equivalent, so that Dynkin systems are λ-systems and vice versa.

A Dynkin system which is also a π-system is a σ-algebra.

Given any collection of subsets of , there exists a unique Dynkin system denoted which is minimal with respect to containing . That is, if is any Dynkin system containing , then . is called the Dynkin system generated by . Note . For another example, let and ; then .

Dynkin's π-λ Theorem

If is a π-system and is a Dynkin system with , then . In other words, the &\sigma;-algebra generated by is contained in .

One application of Dynkin's π-λ theorem is the uniqueness of the Lebesgue measure:

Let (Ω, B, λ) be the unit interval [0,1] with the Lebesgue measure on Borel sets. Let μ be another measure on Ω satisfying μ[(a,b)] = b - a, and let D be the family of sets such that μ[D] = λ[D]. Let I = { (a,b),[a,b),(a,b],[a,b] : 0 < ab < 1 }, and observe that I is closed under finite intersections, that ID, and that B is the σ-algebra generated by I. One easily shows D satisfies the above conditions for a Dynkin-system. From Dynkin's lemma it follows that D is in fact all of B, which is equivalent to showing that the Lebesgue measure is unique.


Bibliography

Gut, Allan (2005). Probability: A Graduate Course. New York: Springer. doi:10.1007/b138932. ISBN 0-387-22833-0.

Billingsley, Patrick (1995). Probability and Measure. New York: John Wiley & Sons, Inc. ISBN 0-471-00710-2.


Dynkin system at PlanetMath.