Product Code Database
Example Keywords: arcade and -jewel $33-149
barcode-scavenger
   » » Wiki: Markov Odometer
Tag Wiki 'Markov Odometer'.
Tag
In mathematics, a Markov odometer is a certain type of topological dynamical system. It plays a fundamental role in and especially in orbit theory of dynamical systems, since a theorem of H. Dye asserts that every nonsingular transformation is orbit-equivalent to a Markov odometer.

The basic example of such system is the "nonsingular odometer", which is an additive topological group defined on the of , induced by addition defined as x \mapsto x+\underline{1}, where \underline{1}:=(1,0,0,\dots). This group can be endowed with the structure of a ; the result is a conservative dynamical system.

The general form, which is called "Markov odometer", can be constructed through Bratteli–Vershik diagram to define Bratteli–Vershik compactum space together with a corresponding transformation.


Nonsingular odometers
Several kinds of non-singular odometers may be defined. These are sometimes referred to as adding machines.
(2025). 9780387304403, Springer.
The simplest is illustrated with the Bernoulli process. This is the set of all infinite strings in two symbols, here denoted by \Omega=\{0,1\}^{\mathbb{N}} endowed with the . This definition extends naturally to a more general odometer defined on the
\Omega=\prod_{n\in\mathbb{N}} \left(\mathbb{Z}/k_n\mathbb{Z}\right)
for some sequence of integers (k_n) with each k_n\ge 2.

The odometer for k_n=2 for all n is termed the dyadic odometer, the von Neumann–Kakutani adding machine or the dyadic adding machine.

The topological entropy of every adding machine is zero. Any continuous map of an interval with a topological entropy of zero is topologically conjugate to an adding machine, when restricted to its action on the topologically invariant transitive set, with periodic orbits removed.


Dyadic odometer
The set of all infinite strings in strings in two symbols \Omega=\{0,1\}^{\mathbb{N}} has a natural topology, the , generated by the . The product topology extends to a Borel ; let \mathcal{B} denote that algebra. Individual points x\in\Omega are denoted as x=(x_1,x_2,x_3,\cdots).

The Bernoulli process is conventionally endowed with a collection of measures, the Bernoulli measures, given by \mu_p(x_n=1)=p and \mu_p(x_n=0)=1-p, for some 0 independent of n. The value of p=1/2 is rather special; it corresponds to the special case of the Haar measure, when \Omega is viewed as a . Note that the Bernoulli measure is not the same as the 2-adic measure on the ! Formally, one can observe that \Omega is also the base space for the dyadic integers; however, the dyadic integers are endowed with a , the p-adic metric, which induces a distinct from the product topology used here.

The space \Omega can be endowed with addition, defined as coordinate addition, with a carry bit. That is, for each coordinate, let (x+y)_n=x_n+y_n+\varepsilon_n\,\bmod\,2 where \varepsilon_0=0 and

\varepsilon_n=\begin{cases} 0 & x_{n-1}+y_{n-1}<2\\ 1 & x_{n-1}+y_{n-1}=2 \end{cases}

inductively. Increment-by-one is then called the (dyadic) odometer. It is the transformation T:\Omega\to\Omega given by T(x)=x+\underline{1}, where \underline{1}:=(1,0,0,\dots). It is called the odometer due to how it looks when it "rolls over": T is the transformation T\left(1,\dots,1,0,x_{k+1},x_{k+2},\dots\right) = \left(0,\dots,0,1,x_{k+1},x_{k+2},\dots \right). Note that T^{-1}(0,0,\cdots)=(1,1,\cdots) and that T is \mathcal{B}-measurable, that is, T^{-1}(\sigma)\in\mathcal{B} for all \sigma\in\mathcal{B}.

The transformation T is non-singular for every \mu_p. Recall that a measurable transformation \tau:\Omega\to\Omega is non-singular when, given \sigma\in\mathcal{B}, one has that \mu(\tau^{-1}\sigma)=0 if and only if \mu(\sigma)=0. In this case, one finds

\frac{d \mu_p \circ T}{d \mu_p} = \left(\frac{1-p} p\right)^\varphi
where \varphi(x)=\min\left\{ n\in\mathbb{N}\mid x_n = 0 \right\}-2. Hence T is nonsingular with respect to \mu_p.

The transformation T is . This follows because, for every x \in \Omega and natural number n, the orbit of x under T^0,T^1,\cdots,T^{2^n-1} is the set \{0,1\}^n. This in turn implies that T is conservative, since every invertible ergodic nonsingular transformation in a is conservative.

Note that for the special case of p=1/2, that \left(\Omega,\mathcal{B},\mu_{1/2},T\right) is a measure-preserving dynamical system.


Integer odometers
The same construction enables to define such a system for every of . In general, one writes
\Omega=\prod_{n\in\mathbb{N}}A_{n}
for A_n=\mathbb{Z}/m_n\mathbb{Z}=\{ 0,1,\dots,m_n-1\} with m_n\ge2 an integer. The product topology extends naturally to the product Borel sigma-algebra \mathcal{B} on \Omega. A on \mathcal{B} is conventionally defined as \textstyle\mu=\prod_{n\in\mathbb{N}}\mu_{n}, given some measure \mu_n on A_n. The corresponding map is defined by
T(x_1,\dots,x_k,x_{k+1},x_{k+2},\dots)=(0,\dots,0,x_k+1,x_{k+1},x_{k+2},\dots)
where k is the smallest index for which x_k \neq m_k-1. This is again a topological group.

A special case of this is the Ornstein odometer, which is defined on the space

\Omega=\left(\mathbb{Z}/2\mathbb{Z}\right)\times \left(\mathbb{Z}/3\mathbb{Z}\right)\times \left(\mathbb{Z}/4\mathbb{Z}\right)\times \cdots
with the measure a product of
\mu_n(j)=\begin{cases}
1/2 & \mbox{ if } j=0 \\ 1/2(n+1) & \mbox{ if } j\ne 0 \\ \end{cases}


Sandpile model
A concept closely related to the conservative odometer is that of the abelian sandpile model. This model replaces the directed linear sequence of finite groups constructed above by an undirected graph (V,E) of vertexes and edges. At each vertex v\in V one places a finite group \mathbb{Z}/n\mathbb{Z} with n=deg(v) the degree of the vertex v. Transition functions are defined by the . That is, one can increment any given vertex by one; when incrementing the largest group element (so that it increments back down to zero), each of the neighboring vertexes are incremented by one.

Sandpile models differ from the above definition of a conservative odometer in three different ways. First, in general, there is no unique vertex singled out as the starting vertex, whereas in the above, the first vertex is the starting vertex; it is the one that is incremented by the transition function. Next, the sandpile models in general use undirected edges, so that the wrapping of the odometer redistributes in all directions. A third difference is that sandpile models are usually not taken on an infinite graph, and that rather, there is one special vertex singled out, the "sink", which absorbs all increments and never wraps. The sink is equivalent to cutting away the infinite parts of an infinite graph, and replacing them by the sink; alternately, as ignoring all changes past that termination point.


Markov odometer
Let B=(V,E) be an ordered Bratteli–Vershik diagram, consists on a set of vertices of the form \textstyle\coprod_{n\in\mathbb{N}}V^{(n)} (disjoint union) where V^0 is a singleton and on a set of edges \textstyle\coprod_{n\in\mathbb{N}}E^{(n)} (disjoint union).

The diagram includes source surjection-mappings s_n:E^{(n)} \to V^{(n-1)} and range surjection-mappings r_n:E^{(n)} \to V^{(n)}. We assume that e,e' \in E^{(n)} are comparable if and only if r_n(e) = r_n(e').

For such diagram we look at the product space \textstyle E:=\prod_{n\in\mathbb{N}}E^{(n)} equipped with the . Define "Bratteli–Vershik compactum" to be the subspace of infinite paths,

X_{B}:=\left\{ x=(x_n)_{n\in\mathbb{N}} \in E\mid x_n\in E^{(n)}\text{ and } r (x_n) = s(x_{n+1}) \right\}

Assume there exists only one infinite path x_{\max} = (x_n)_{n \in \mathbb{N}} for which each x_n is maximal and similarly one infinite path x_{\text{min}}. Define the "Bratteli-Vershik map" T_B:X_B \to X_B by T( x_{\max}) = x_{\min} and, for any x = (x_n)_{n\in \mathbb{N}} \neq x_{\max} define T_B(x_1,\dots,x_k,x_{k+1},\dots)=(y_1,\dots,y_k,x_{k+1},\dots), where k is the first index for which x_k is not maximal and accordingly let (y_1,\dots,y_k) be the unique path for which y_1,\dots,y_{k-1} are all maximal and y_k is the successor of x_k. Then T_B is of X_B.

Let P=\left(P^{(1)},P^{(2)},\dots \right) be a sequence of stochastic matrices P^{(n)}=\left(p^{(n)}_{(v,e) \in V^{n-1} \times E^(n)}\right) such that p^{(n)}_{v,e} > 0 if and only if v=s_n(e). Define "Markov measure" on the cylinders of X_B by \mu_P (e_1,\dots,e_n) = p^{(1)}_{s_1(e_1),e_1}\cdots p^{(n)}_{s_n(e_n),e_n}. Then the system \left(X_B, \mathcal{B}, \mu_P, T_B \right) is called a "Markov odometer".

One can show that the nonsingular odometer is a Markov odometer where all the V^{(n)} are singletons.


See also
  • Abelian sandpile model


Further reading

Page 1 of 1
1
Page 1 of 1
1

Account

Social:
Pages:  ..   .. 
Items:  .. 

Navigation

General: Atom Feed Atom Feed  .. 
Help:  ..   .. 
Category:  ..   .. 
Media:  ..   .. 
Posts:  ..   ..   .. 

Statistics

Page:  .. 
Summary:  .. 
1 Tags
10/10 Page Rank
5 Page Refs
1s Time