Steve the presenter
Download
1 / 36

Cayley Graphs Expanders - PowerPoint PPT Presentation


  • 129 Views
  • Uploaded on

Steve the presenter. Cayley Graphs & Expanders. Some history. Arthur Cayley. Son of a British merchant in Russia who wedded a Russian girl. English Super Genius Did everything Very smart and knew more than 6 languages including yours and math.

loader
I am the owner, or an agent authorized to act on behalf of the owner, of the copyrighted work described.
capcha
Download Presentation

PowerPoint Slideshow about 'Cayley Graphs Expanders' - glora


An Image/Link below is provided (as is) to download presentation

Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.


- - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - -
Presentation Transcript
Steve the presenter

Steve the presenter

Cayley Graphs & Expanders


Some history
Some history

Arthur Cayley


The Real Arthur Cayley

For more demoralizing facts, visit his homepage on Wikipedia


Some definitions
Some Definitions girl.

Let G be a graph (finite and infinite), and let S be a non-empty, finite subset of G. Here we assume S to be symmetric, S= S −1

Definition: The Cayley graph G(G, S) is the graph with the vertex set V = G and edge set

E = {{x,y} : x,y in G, there exist s in S : y = xs}


  • Hence, two vertices are girl. adjacent if one is obtained from the other by right multiplication by some element of S.

  • Note, since S is symmetric, this adjacency relation is also symmetric, i.e, a ~ b implies b ~ a, so that the resulting graphs are undirected.



Let s look at some examples
Let's look at some examples: girl.

If G = Zn is the finite cyclic group of order n and the set S consists of two elements, the standard generator of G and its inverse, then the Cayley graph is the cycle Cn.


G z 6 z s 1 1
G girl. = Z/6Z, S = {1, -1}










More examples
MOre Examples girl.

Similarly, suppose now that G= Z is the infinite cyclic group and the set S consists of the standard generator 1 and its inverse (−1 in the additive notation) then the Cayley graph is an infinite chain.


G z s 1 1
G girl. = Z, S = {1, -1}


G z s 1 11
G girl. = Z, S = {1, -1}

This of course simply the number line.


G z s 2 2 3 3
G girl. = Z, S = {2, -2, 3, -3}


G z 2 s 0 1 1 0 0 1 0 1
G girl. = Z2, S = {(0,1), (-1,0), (0,1), (0,-1)}


G l2 the free group on 2 generators a b s a a 1 b b 1
G girl. = L2, the free group on 2 generators a, b; S = {a, a-1, b, b-1}


Alternatively, if we draw each new edge horizontal and vertical to half the length, this gives rise to fractal images. The Cayley graphs of infinite groups provide interesting geometries!


Expanders
                     Expanders vertical to half the length, this gives rise to fractal images. The Cayley graphs of infinite groups provide interesting geometries!

Now what is an expander?


The investigation of an expander centres around the following question:

Given a sparse set of points, how do we systematically construct a highly connected yet efficient network between the points?

Conceivably, the answer to such a question is of great interest to electric grid constructors and designers of network between computers among many others.


Cayley graphs groups and expanders
Cayley Graphs, Groups and expanders following question:

In the previous slides, we have seen that Cayley graphs give very nice encoding of the structure of discrete groups.

Conversely, it also helps us sometimes to translate a graph problem back into its group presentation. This way, we can I will not tell you how now because I don't know yet. But do remember to ask me about it if you see me next semester!


Expanders1
Expanders following question:

We consider graphs X = (V, E), where V again is the set of vertices and E the set of edges of X. Here we shall limit ourselves to the case that X is undirected and finite.


  • A following question: path in X is a sequence v1, v2,...., vk of vertices, where vi is adjacent to vi+1 (i.e. and {vi, vi+1} is an edge).

  • A graph X is connected if every two vertices can be joined by a path. (One can trace from one vertice to another one through a series of edges.)‏

  • Finally a graph is k-regular if every vertex is also connected to k other vertices through k-edges.


  • This is the celebrated following question: Petersen Graph.

  • It is clearly connected.

  • In fact, it is a 3-regular expander.


  • Let F be a subset of following question: V, The boundary ∂F is the the set of edges connecting F to V – F.

  • In our case here, F = 3, V – F = 7, ∂F are the seven black “fat” edges.



  • h(X) = inf {abs(∂F)/(min{abs(F),abs(V – F)}) : F is subset of V, 0 < abs(F) < +∞}

  • If we view X as a network transmitting information (where retained by some vertex propagates, say in one unit of time, to neighboring vertices), then h(X) measures the quality of X as a network:

  • h(X) is large, information propagates well.


Let s consider two extreme examples
Let's consider two extreme examples: subset of V, 0 < abs(F) < +∞}

  • The complete graph Km on m vertices is defined s.t. Every vertex is connected to one another, e.g. m = 5.

  • It's clear in this case, if F = i, ∂F = i(m-i) s.t. h(Km) = m/2


  • The cycle Cn on n vertices: subset of V, 0 < abs(F) < +∞}

  • When n = 6, if F is a half-cycle, ∂F = 2

  • h(Cn) is smaller than 2/(n/2) = 4/n

  • In particular, h(Cn) goes to zero as n goes to infinity


  • Thus we see that highly connected complete graph has a large expanding constant that grows proportionately with the number of vertices.

  • On the other hand, the minimally connected graph has a small expanding constant that goes to zero as n increases.

  • In this sense, h(X) does indeed give us a measure of the “quality” of the network.


That's all I know about expanders at this point. expanding constant that grows proportionately with the number of vertices.

Thank you!


ad