Lec 30

Download as pdf or txt
Download as pdf or txt
You are on page 1of 12

Preferential Attachment Model

Addition/Deletion of Edges
MA 653: Network Science
Instructor: Ashok Singh Sairam
ashok@iitg.ac.in
Extensions of preferential attachment
• Price and BA models include many oversimplified assumptions for the
way networks grow
• In particular, they assume that nodes and edges are only created (never
deleted) and a vertex can initiate new edges only at the time of creation
• While these assumptions might hold with good approximation for
citation networks they do not hold true for other types of networks
• In Web links are not permanent and can be created at any time
• Entire webpages can also disappear (vertex removal)
• Why the preferential attachment should be linear in the degree?

MA 653: Network Science 2


Addition of extra edges
• While in citation networks no new edges can be created by a vertex
after the time of the node generation this is not true for the majority
of the networks
• For example, the WWW is constantly changing
• A network evolution model that can deal with the creation of new
edges is the following:
• At each step, the new node creates exactly c new edges which attach to other
vertices with probability proportional to degree k [same as before]
• In addition, at each step some number w of extra edges are added to the
network with both ends attaching to vertices chosen in proportional to
degree
• Thus, when the network has n vertices it will have n(c+w) edges

MA 653: Network Science 3


Addition of extra edges
• Every new node brings c+2w ends of edges
• 2 extra for each of the w extra edges
• The normalization factor for the attaching probabilities is the sum of
the degree of all nodes
• This is equal to twice the number of edges, that is, 2n(c+w)
• The probability that a node with degree k will attract one of the new
edges is:
k c + 2w
npk (n)´ (c + 2w)´ = kpk (n)
2n(c + w) 2(c + w)
• pk(n) is the fraction of nodes with degree k when the network has n vertices

MA 653: Network Science 4


Addition of extra edges
• We can then write the master equation
c + 2w
(n +1)pk (n +1) = npk (n) + [(k -1)pk-1 (n) - kpk (n)], k > c
2(c + w)
c + 2w
(n +1)pc (n +1) = npc (n) +1- cpc (n), k=c
2(c + w)
• Taking the limit of large n and solving the equations we get:
𝐵(𝑘, 𝛼) 𝑐
𝑝𝑘 = where 𝛼 = 2 +
𝐵(𝑐, 𝛼 − 1) 𝑐 + 2𝑤
• Again if we take the asymptotic behavior of Beta function we see that the degree
distribution has a power law tail with exponent a
• For w=0 we get 𝛼 =3 as we should have expected (BA model)
• For w>0, we get 2< 𝛼 <3
MA 653: Network Science 5
Removal of edges
• Edges can be removed as well in many networks
• Here we begin with an extension of the original BA model to account for
deletion of edges at each step (edges additions are still only happening at the
initial creation of a vertex)
• In particular, at each step a new vertex is created with c edges associated with
him
• These edges are attached to existing nodes with the original BA process,
that is, proportional to the existing nodes degrees
• At each step we also remove u edges from the existing ones
• The edges that are deleted are picked uniformly at random
2ki
• The probability that a node of degree ki loses one of its edges is:
åk i i

MA 653: Network Science 6


Removal of edges
• For edges in the network to grow and not shrink with time it needs to hold: u<c
• The total number of edges after n steps is n(c-u)
• When writing the master equation we need to be careful since there are more
ways now to get nodes of degree k
• The expected number of vertices of degree k that will get a new edge at the
(n+1)th step is:
𝑘 𝑐
𝑛𝑝𝑘 𝑛 𝑐 = 𝑘𝑝𝑘 (𝑛)
2𝑛(𝑐 − 𝑢) 2 𝑐 − 𝑢
• The expected number of vertices of degree k that will lose one edge at the
(n+1)th step is:
2k u
npk ( n )  u   kpk ( n )
2n ( c  u ) c  u
MA 653: Network Science 7
Removal of edges
• As it should be evident nodes now can have degrees less than c as well
• The master equation takes the following form:
c u c u
(n +1)pk (n +1) = npk (n) + (k -1)pk-1 (n) + (k +1)pk+1 (n) - kpk (n) - kpk (n), k ¹ c
2(c - u) c-u 2(c - u) c-u
c u c u
(n +1)pc (n +1) = npc (n) +1+ (c -1)pc-1 (n) + (c +1)pc+1 (n) - cpc (n) - cpc (n), k = c
2(c - u) c-u 2(c - u) c-u

• These equations can be combined to:


c 2u c + 2u
(n +1)pk (n +1) = npk (n) + dkc + (k -1)pk-1 (n) + (k +1)pk+1 (n) - kpk (n)
2(c - u) 2(c - u) 2(c - u)
• Problem arises for k=0, since it involves p-1(n). To overcome this problem we define p-1=0

MA 653: Network Science 8


Addition/Removal of edges
• This extension of BA model accounts only for removal of edges, while
it allows creation of edges only during the initial creation of a vertex
• We can easily add to the above process the addition of w edges (on
top of the c edges created by the new vertex) at every node creation
instance
• The master equation takes the form:
c + 2w u c + 2w + 2u
(n +1)pk (n +1) = npk (n) + dkc + (k -1)pk-1 (n) + (k +1)pk+1 (n) - kpk (n)
2(c + w - u) c+w-u 2(c + w - u)
• In this case we require u<w+c

MA 653: Network Science 9


Addition/Removal of edges
• Taking the limit of large n we can obtain the equations for pk
• However, as we can observe they are different in some aspect from what we
have seen until now
• In particular, they involve probabilities for 3 degrees
• k-1, k and k+1
• The presence of three different degree terms makes it substantially
more difficult to solve
• We need to make use of the generating functions and eventually we obtain a
first-order linear differential equation

MA 653: Network Science 10


Addition/Removal of edges
• The final result for a large k (and u<w+(1/2)c) is:
G(a -1) -a u-w 2u
pk ~ k , a = 2 + g =
(1- g )a-1 c + 2w - 2u c + 2w
• As we can see this extension of the model provides a power law tail
again
• The exponent can take values both greater and smaller than 2
• However, for u=w+(1/2)c, the exponent becomes infinite
• At this point we lose the power law behavior and the distribution becomes a stretched
exponential
• Also for u >w+(1/2)c, the solution becomes nonsensical and we need to solve
the differential equation with other methods

MA 653: Network Science 11


Exercise
Ex. 13.5: Consider the following variant of the Barabási-Albert model. Nodes are
added one-by-one to a growing undirected network. Each node having initial
degree c. The c edges emanating from a newly added node connect to previously
existing nodes i with probability ki+a.
a) Given that c edges are added to the network with each node, what is the mean
degree of a node in the network in the limit of large network size?
b) Derive the master equation that gives the fraction of nodes pk having degree k
in the limit of large network size. If necessary give an additional rate equation
to cover any special-case value of k.

MA 653: Network Science 12

You might also like