ArticlePDF Available

Lumping the Approximate Master Equation for Multistate Processes on Complex Networks


Abstract and Figures

Complex networks play an important role in human society and in nature. Stochastic multistate processes provide a powerful framework to model a variety of emerging phenomena such as the dynamics of an epidemic or the spreading of information on complex networks. In recent years, mean-field type approximations gained widespread attention as a tool to analyze and understand complex network dynamics. They reduce the model's complexity by assuming that all nodes with a similar local structure behave identically. Among these methods the approximate master equation (AME) provides the most accurate description of complex networks' dynamics by considering the whole neighborhood of a node. The size of a typical network though renders the numerical solution of multistate AME infeasible. Here, we propose an efficient approach for the numerical solution of the AME that exploits similarities between the differential equations of structurally similar groups of nodes. We cluster a large number of similar equations together and solve only a single lumped equation per cluster. Our method allows the application of the AME to real-world networks, while preserving its accuracy in computing estimates of global network properties, such as the fraction of nodes in a state at a given time.
Content may be subject to copyright.
Lumping the Approximate Master Equation for
Multistate Processes on Complex Networks
Gerrit Großmann1, Charalampos Kyriakopoulos1, Luca Bortolussi2, and
Verena Wolf1
1Computer Science Department, Saarland University
2Department of Mathematics and Geosciences, University of Trieste
Abstract. Complex networks play an important role in human society
and in nature. Stochastic multistate processes provide a powerful frame-
work to model a variety of emerging phenomena such as the dynamics of
an epidemic or the spreading of information on complex networks. In re-
cent years, mean-field type approximations gained widespread attention
as a tool to analyze and understand complex network dynamics. They
reduce the model’s complexity by assuming that all nodes with a similar
local structure behave identically. Among these methods the approxi-
mate master equation (AME) provides the most accurate description of
complex networks’ dynamics by considering the whole neighborhood of a
node. The size of a typical network though renders the numerical solution
of multistate AME infeasible. Here, we propose an efficient approach for
the numerical solution of the AME that exploits similarities between the
differential equations of structurally similar groups of nodes. We clus-
ter a large number of similar equations together and solve only a single
lumped equation per cluster. Our method allows the application of the
AME to real-world networks, while preserving its accuracy in computing
estimates of global network properties, such as the fraction of nodes in
a state at a given time.
Keywords: Complex Networks, Multistate Processes, AME, Model Re-
duction, Lumping
1 Introduction
Various emerging phenomena of social, biological, technical, or economic nature
can be modeled as stochastic multistate processes on complex networks [1, 3,
24, 26]. Such networks typically consist of millions or even billions of nodes [1,
3], each one being in one of a finite number of states. The state of a node can
potentially change over time as a result of interaction with one of its neighbor-
ing nodes. The interactions among neighbors are specified by rules and occur
independently at random time points, governed by the exponential distribution.
Hence, the underlying process is a discrete-state space Markovian process in
continuous time (CTMC). Its state space consists of all labeled graphs repre-
senting all possible configurations of the complex network. For instance, in the
arXiv:1804.02981v1 [cs.SI] 9 Apr 2018
susceptible-infective (SI) model, which describes the spread of a simple epidemic
process, each node can either be susceptible or infected; infected nodes propagate
the infection to their susceptible neighbors [19, 5].
Monte-Carlo simulations can be carried out only for small networks [11, 19],
as they become very expensive for large networks, due to the large number
of simulation runs which are necessary to draw reliable conclusions about the
network’s dynamics.
An alternative and viable approach is based on mean-field approximations, in
which nodes sharing a similar local structure are assumed to behave identically
and can be described by a single equation, capturing their mean behavior [18,
3, 4, 10, 12]. The heterogeneous (also called degree-based) mean-field (DBMF)
approach proposes a system of ordinary differential equations (ODEs) with one
equation approximating the nodes of degree kwhich are in a certain state [25,
9, 18]. The approximate master equation (AME) provides a far more accurate
approximation of the network’s dynamics, considering explicitly the complete
neighborhood of a node in a certain state [16, 17, 14]. However, the corresponding
number of differential equations that have to be solved is of the order Ok|S|
where kmax is the network’s largest degree and |S| the number of possible states.
A coarser approximation called pair approximation (PA) can be derived from
AME by imposing the multinomial assumption for the number of neighbors in
a state [16, 17]. Nevertheless, solving PA instead of AME is faster but for many
networks not accurate enough [17].
Lumping is a popular model reduction technique for Markov-chains and sys-
tems of ODEs [21, 6, 28, 7, 8]. It has also been applied to the underlying model
of epidemic contact processes [27, 19] and has recently been shown to be ex-
tremely effective for the DBMF equation as well as for the PA approach [20].
In this work, we generalize the approach of [20] providing a lumping scheme for
the AME, leveraging the observation that nodes with a large degree having a
similar neighborhood structure have also typically very similar behaviors. We
show that it is possible to massively reduce the number of equations of the AME
while preserving the accuracy of global statistical properties of the network. Our
contributions, in particular, are the following: (i) we provide a fully automated
aggregation scheme for the multistate AME; (ii) we introduce a heuristic to find
a reasonable trade-off between number of equations and accuracy; (iii) we evalu-
ate our method on different models from literature and compare our results with
the original AME and Monte-Carlo simulation; (iv) we provide an open-source
tool3written in Python, which takes as input a model specification, generates
and solves the lumped (or original) AME.
The remainder of this paper is organized as follows: In Section 2 we describe
multistate Markovian processes in networks and formally introduce the AME. In
Section 3 we derive lumped equations for a given clustering scheme and in Section
4 we propose and evaluate a clustering algorithm for grouping similar equations
together. Case studies are presented in Section 5. We draw final conclusions and
identify open research problems in Section 6.
2 The Multistate Approximate Master Equation
In this section, we first define contact processes and introduce our notation and
terminology for the multistate AME.
2.1 Multistate Markovian Processes
We describe a contact process in a network (G,S, R, L) by a finite undirected
graph G= (V, E ), a finite set of states S, a set of rules R, and an initial state
for each agent (node) of the graph L:V→ S. We use s, s0, s00 and s1, s2, . . . to
denote elements of S. At each time point t0, each node vVis in a state
s∈ S. The rules Rdefine how neighboring nodes influence the probability of
state transitions. A rule consists of a consumed state, a produced state, and a
transition rate, which depends on the neighborhood of the node. We use integer
vectors to model a node’s neighborhood. For a given set of states Sand maximal
degree kmax, the set of all potential neighborhood vectors is M={mZ|S|
Ps∈S m[s] kmax}, where we write m[s] to refer to the number of neighbors in
state s.
A rule rRis a triplet r= (s, f, s0) with s, s0∈ S , s 6=s0and rate function
f:M → R0corresponding to the exponential distribution. A rule r(also
denoted as sf
s0) can be applied at every node in state s, and, when applied,
it transforms this node into state s0. Note that this general formulation of a rule
containing the rate function can express all types of rules that are described in
[16, 17, 14] such as spontaneous changes of a node’s state (independent rules) or
changes due to the state of a neighbor (contact rules). The delay until a certain
rule is applied is exponentially distributed with rate f(m), with rules competing
in a race condition where the one with the shortest delay is executed. This results
in an underlying stochastic model described by a CTMC.
In the following, we indicate with Rs+={(s0, f, s)R, s0∈ S} all the rules
that change the state of a node into s, and with Rs={(s, f, s0)R, s0∈ S}
all rules that change an s-node into a different state.
Example In the SIS model, a susceptible node can become infected by one
of its neighbors. An infected node becomes susceptible again, independently of
its neighbors. Hence, the infection rule is S λ1·m[I]
I and the recovery rule is
S,where m[I] denotes the number of infected neighbors and λ1, λ2R0
are rule-specific rate constants.
2.2 Multistate AME
Here, we briefly present the multistate AME, similarly to [14, 20]. The AME
assumes that all nodes in a certain state and with the same neighborhood struc-
ture are indistinguishable. We define Mk={m∈ M | Ps∈S m[s] = k}to be
the subset of neighborhood vectors referring to nodes of degree k. In addition,
for s1, s2∈ S and m∈ M, we use m{s
2}to denote a neighborhood vector
where all entries are equal to those of m, apart from the s1-th entry, which is
equal to m[s1] + 1, and the s2-th entry, which is equal to m[s2]1.
Let xs,m(t) be the fraction of network nodes that are in state sand have
a neighborhood mat time t, and assume the initial state xs,m(0) is known.
Formally, the AME approximates the time evolution of xs,mwith the following
set of deterministic ODEs4:
∂t =X
where, the term βss1ss2is the the average rate at which an (s, s1)-edge changes
into an (s, s2)-edge, if s, s1, s2∈ S with s16=s2.
The first term in the right hand side models the inflow into (s, m) nodes
from (s0,m) nodes, while the second term models the outflow from (s, m) due
to the application of a rule. The other two terms describe indirect effects on a
(s, m) node due to changes in its neighboring nodes, again considering inflow
and outflow (cf. Fig. 1). In particular, a node in the neighborhood mof (s, m),
say in state s1, changes to state s2by the firing of a rule.
To compute βss1ss2we need to define the subset of rules which consume a
s1-node and produce an s2-node: Rs1s2={(s1, f, s2)R|f:M → R0}.
m∈M P
where in the denominator we normalize dividing by the fraction of (s, s1) edges.
The total number of equations of AME is determined by the number of states
|S| and the maximal degree kmax , and equals:
kmax +|S|
|S| − 1(kmax + 1) .(3)
The binomial arises from the number of ways in which, for a fixed degree k, one
can distribute kneighbors into |S| different states, see [20] for the proof.
As xs,mare fractions of network nodes, the following identity holds for all t:
xs,m(t) = 1 (4)
4we omit tfor the ease of notation
node changes neighborhood changes
Fig. 1: Illustration of how the AME governs the fraction of xs,(2,2) in a two-state
model with rules (s0, f1, s), (s, f2, s0). The inflow and outflow between xs,(2,2)
and xs0,(2,2) is induced by the direct change of a node’s state from sto s0or vice
versa. The inflow and outflow between xs,(2,2) and xs,(3,1),xs,(1,3) is attributed
to the change of state of a node’s neighbor.
Moreover, we use xsto denote the global fraction of nodes in a fixed state s,
which we get by summing over all possible neighborhood vectors
xs(t) = X
again with Ps∈S xs(t) = 1. Intuitively, xsis the probability that a randomly
chosen node from the network is in state s. This is the value of primary interest
in many applications, e.g. [3, 24, 26]. Finally, the degree distribution P(k) gives
the probability that a randomly chosen node is of degree k(0 kkmax). If
we sum up all xs,mwhich belong to a specific k(i.e. m∈ Mk), as the network
structure is assumed to be static, we will necessarily obtain the corresponding
degree probability. Hence, for each t0, we have
xs,m(t) = P(k).(6)
3 Lumping
The key idea of this paper is to group together equations of the AME which have
a similar structure and to solve only a single lumped equation per group. This
lumped equation will capture the evolution of the sum of the AME variables in
each group.
Therefore, we divide the set {xs,m|s∈ S,m∈ M} into groups or clusters,
constructing our clustering such that two equations xs,m,xs0,m0can only end
up in the same group if s=s0and mis ‘sufficiently’ similar to m0. This ensures
that the fractions within a cluster as well as their time derivatives are similar,
provided the change in the rate as a function of mis relatively small when mis
In the sequel, we consider a clustering Cdefined as a partition over M, i.e.,
C ⊂ 2Mand SC∈C C=Mand all clusters Care disjoint and non-empty. Before
we discuss in detail the construction of Cin Section 4, we derive the lumped
equations for a given clustering C.
First, recall that we want to approximate the global fractions for each state
(cf. Eq. (5)), which can be split into sums over the clusters
xs(t) = X
Our goal is now to construct a smaller equation system, where the variables
zs,C approximate the sum over all xs,mwith mC
zs,C (t)X
Henceforth, we can approximate the global fractions as
zs,C (t).(9)
The number of equations is then given by |S| · |C |. As one might expect, there is
a trade-off between the accuracy of zs,C(t) and the computational cost, propor-
tional to the number of clusters.
3.1 Lumping the Initial State and the Time Derivative
As the initial values of xs,mare given, we define the initial lumped values
zs,C (0) = X
xs,m(0) .(10)
To achieve the criterion in Eq. (8) for the fractions computed at t > 0, we
seek for time derivatives which fulfill
∂t X
∂t .(11)
Note that an exact lumping is in general not possible as ∂zs,C
∂t is a function of the
individual xs,m(t). In order to close the equations for zs,C , we need to express
xs,mas an approximate function of zs,C. The naive idea is to assume that the
true fractions xs,mare similar for all mthat belong to the same cluster, i.e., if
m,m0Cthen xs,mxs,m0, leading to an approximation of xs,mas zs,C/|C|.
This is however problematic, as it neglects the fact that neighbors of nodes
of different degree have different size. In fact, even if for two degrees k1< k2
in the same cluster we have P(k1) = P(k2) (while typically P(k1)> P (k2)),
the number of possible different neighbors m2of a k2-node is larger than the
number of different neighbors m1of a k1-node, |Mk1|<|Mk2|, hence typically
xs,m2< xs,m1, as the mass of P(k2) has to be split among more variables. In
order to correct for this asymmetry between degrees in each cluster, we introduce
the following assumption:
Assumption:All fractions xs,minside a cluster Cthat refer to the same degree
contribute equally to the sum zs,C . Equations of different degree contribute pro-
portionally to their degree probability P(k)and inversely proportionally to the
neighborhood size for that degree.
Based on the above assumption, we define a degree dependent scaling-factor
wC,k R0, which only depends on the corresponding cluster Cand degree k.
According to the above assumption wC,k P(k)
|Mk|. To ensure that the weights of
one cluster sum up to one, we define
wC,k =P(k)
|Mk|· X
where km=Ps∈S m[s] is the degree of a neighborhood m. We compute approx-
imations of xs,mbased on zs,C as
xs,mzs,C ·wC,km.(13)
3.2 Building the Lumped Equations
To define a differential equation for the lumped fraction zs,C , we consider again
Eq. (11) and replace xs,m
∂t by the l.h.s. of Eq. (1). Then we substitute every
occurrence of xs,mby its corresponding lumped variable multiplied with the
scaling factor, i.e., zs,C ·wC,km,where mC. Since mCdoes generally
not imply that m{s
2}C, the substitution of xs,m{s
2}is somewhat more
complicated. Let C(m) denote the cluster mbelongs to. If mlies “at the border”
of a cluster then C(m{s+
2}) might be different than C(m). The lumped AME
takes then the following form:
∂t =X
zs0,C X
zs,C X
Lzs,C X
zs1,C P
zs1,C P
To gain a significant speedup compared to the original equation system, it is
necessary that the lumped equations can be efficiently evaluated. In particular,
we want the number of terms in the lumped equation system to be proportional
to the number of fractions zs,C and not to the number of xs,m. This is possi-
ble for Eq. (14), because each time we have a sum over mC, for instance
PmCwC,kmf(m), we can precompute this value during the generation of the
equations and do not have to evaluate it at every step of the ODE solver. The
can be evaluated efficiently since we only have to consider lumped variables
that correspond to clusters C(m{s
2}) that are close to C(m), i.e., that can
be reached from a state in C(m) by the application of a rule. The number of
such neighboring clusters is typically small, due to our definition of clusters, see
Section 4.
Remark 1. For large kmax, the number of neighbor vectors in M, i.e. the size of
the AME, becomes prohibitively large. For instance, for a maximum degree of
the order of 10 thousands, quite common in real networks, the size of Mbecomes
of the order of 1012. Even summing a number of elements of this order while
generating equations becomes very costly. To overcome this limit, the solution is
to approximate terms involving summations in Eq. (14). Consider for instance
PmCwC,kmf(m). Instead of evaluating fat every mCand averaging it
w.r.t. wC,km, we can only evaluate fat the mean neighborhood vector hmiC,
where each coordinate is defined as hmiC[s] = PmCwC,kmm[s]. We can then
approximate PmCwC,kmf(m)fhmiC. See Appendix A for details.
4 Partitioning of the Neighborhood Set
In this section we describe an algorithm to partition M, and construct the clus-
tering C. Our algorithm builds partitions with a varying granularity to control
the trade-off between accuracy and execution speed. We consider three main
criteria: the similarity of different equations, their impact on the global error,
and how fast is the evaluation of the lumped equations. Furthermore, as the size
of Mcan be extremely large, we cannot rely on typical hierarchical clustering
algorithms having a cubic runtime in the number of elements to be clustered.
Our solution is to decouple each minto two components: its degree km(encod-
ing its length) and its projection to the unit simplex (encoding its direction).
We cluster these two components independently.
4.1 Hierarchical Clustering for Degrees
Since our clustering is degree-dependent, we first partition the set of degrees
{0, . . . , kmax}. Let K ⊂ 2{0,...,kmax }be a degree partitioning, i.e., the disjoint
union of all K∈ K is the set of degrees. The goal of the degree clustering is to
merge together consecutive degrees with small probability while putting degrees
with high probability mainly in separate clusters. This is particularly relevant for
the power-law distribution, which is predominantly found in real world networks
[2, 1] as it allows us to cluster a large number of high degrees with low total
probability all together without losing much information.
We use an iterative procedure inspired by bottom-up hierarchical clustering
to determine K. We start by assigning to each degree an individual cluster and
iteratively join the two consecutive clusters that increase the cost function L
by the least amount. The cost function Lpunishes disparity in the spread of
probability mass over clusters, leading to clusters that have approximately the
same total probability mass. It is defined as
L(K) = X
Note that L(K) is minimal when all PkKP(k) have equal values. The algorithm
needs O(k2
max) comparisons to determine the degree cluster of each element. At
the end of this procedure, each m∈ M has a corresponding degree-cluster K
with kmK.
4.2 Proportionality Clustering
Independently of K, we partition Malong the different components of vectors
m∈ M. First, observe that if we normalize mby dividing each dimension by
km, we can embed each Mkinto the unit simplex in R|S| . The idea is then
to partition the unit simplex, and apply the same partition to all Mk. More
specifically, we construct such partition coordinate-wise. As each element of the
normalized mtakes values in [0,1], we split the unit interval in p+1 subintervals
(a) (b)
Fig. 2: Left: Clustering of Mfor a 2–state model with kmax = 20 and |K| =
|P| = 7. Right: Proportionality cluster of a 3–state model with kmax = 50 and
|P| = 5. Only the plane M50 is shown.
p,1]}. Then, two normalized neighbor vectors are in
the same proportionality cluster if and only if their coordinates all belong to the
subinterval P∈ P, possibly different for each coordinate.
4.3 Joint Clusters
Finally, we construct Csuch that two points m,m0are in the same cluster if
and only if they are in the same degree-cluster (i.e., K∈ K :km, km0K) and
in the same proportionality cluster, (i.e., for each dimension s∈ S, there exists
aP∈ P, such that m[s]
The effect of combining degree and proportionality clusters, for a model with
two different states, is shown in Fig. 2a, where the proportionality clustering gives
equally sized triangles that are cut at different degrees by the degree clustering. If
we fix a degree k, each cluster has only two neighbors (one in each direction). In
the 3d-case, the proportionality clustering creates tetrahedra, which correspond
to triangles if we fix a degree k(cf. Fig. 2b).
The above clustering admits some advantageous properties: (1) If we fix a
degree, all clusters have approximately the same size and spatial shape; (2) The
number of ‘direct spatial neighboring clusters’ of each cluster is always small,
which simplifies the identification of clusters in the ‘border’ cases and eases
the generation and evaluation of the lumped AME. Hence, the clusters can be
efficiently computed even if Mis very large. Next, we discuss how to choose the
size of the clusterings Kand P.
4.4 Stopping Heuristic
To find an adequate number of clusters, we solve the lumped AME of the model
multiple times while increasing the number of clusters. We stop when the differ-
ence between different lumped solutions converges. The underlying assumption
is that the approximations become more accurate with an increasing number
of clusters and that the respective difference between consecutive lumped solu-
tions becomes evidently smaller when the error starts to level off. Our goal is to
stop when the increase in the number of clusters does not bring an appreciable
increase in accuracy.
Let z0(t), z00(t) be two solution vectors, i.e., containing the fractions of nodes
in each state at time t, of the lumped AME that correspond to two different
clusterings C0and C00.We define the difference between two such solutions z0,
z00 as their maximal Euclidean distance over time.
(z0,z00) = max
s∈S z0
For the initial clustering we choose |K| =|P| =c0. In each step, we increase the
number of clusters by multiplying the previous ciwith a fixed constant, thus
ci+1 =brcic(r > 1). We find this to be a more robust approach than increasing
ciby only a fixed amount in each step. We stop when the difference between
two consecutive solutions are smaller than stop >0. We consistently observe in
all our case studies that (z0,z00) is a very good indicator on the behavior of the
real error (cf. Fig. 6b). For our experiments we set empirically c0= 10, r= 1.3,
and stop = 0.01.
5 Case Studies
We demonstrate our approach on three different processes, namely the well-
known SIR model, a rumor spreading model, and a SIS model with competing
pathogens [22, 13]. We test how the number of clusters, the accuracy, and the
runtime of our lumping method relate. In addition, we compare the dynamics
of the original and lumped AME with the outcome of Monte-Carlo simulations
on a synthetic network of 105nodes [15, 23]. We performed our experiments
on an Ubuntu machine with 8 GB of RAM and quad-core AMD Athlon II X4
620 processor. The code is written in Python 3.5 using SciPy’s vode5ODE
solver. The lumping error we provide is the difference between lumped solutions
(corresponding to different granularities) and the outcome of the original AME.
That is, for the original solution xand a lumped solution z, we define the
lumping errors of zas (x,z). To generate the error curves, we start with |P| =
|K| = 5 and increase both quantities by one in each step. Note that we test our
approach on models with comparably small kmax. In general, this undermines
the effectiveness of our lumping approach; however using a larger kmax would
have hindered the generation of the complete error curve.
(a) (b)
Fig. 3: SIR model. (a): Lumping error and runtime of the ODE solver w.r.t. the
number of clusters. (b): Fractions of S,I,R nodes over time, as predicted by the
original AME (solid line), by the lumped AME (dashed line), and based on
Monte-Carlo simulations (diamonds).
5.1 SIR
First, we examine the well-known SIR model, where infected nodes (I) go through
a recovery state (R) before they become susceptible (S) again:
I I λ2
R R λ3
We choose (λ1, λ2, λ3) = (3.0,2.0,1.0) and assume a network structure with
kmax = 60 and a truncated power-law degree distribution with γ= 2.5. The
initial distribution is (xI(0), xR(0), xS(0)) = (0.25,0.25,0.5).
In this model the lumping is extremely accurate. In particular, we see that the
lumping error of our method becomes quickly very small (Fig. 3a) and that we
only need a few hundred ODEs to get a reasonable approximation of the original
AME. The lumped solution zwe get from the stopping heuristic, consisting of less
than 5% of the original equations, is almost indistinguishable from the original
AME solution xand the Monte-Carlo simulation (Fig. 3b). The lumping error is
(x,z)=0.0015.The lumped solution used here 1791 clusters with a runtime of
235 seconds while solving the original AME we needed 39711 clusters and 7848
5.2 Rumor Spreading
In the rumor spreading model [13], agents are either ignorants (I) who do not
know about the rumor, spreaders (S) who spread the rumor, or stiflers (R)
who know about the rumor, but are not interested in spreading it. Ignorants
learn about the rumor from spreaders and spreaders lose interest in the rumor
when they meet stiflers or other spreaders. Thus, the rules of the model are the
S S λ2·m[R]
R S λ3·m[S]
(a) (b)
Fig. 4: Rumor spreading model. (a): Lumping error and runtime of the ODE
solver w.r.t. the number of clusters. (b): Fractions of nodes in each state over
time given by the original AME (solid line), by the lumped AME (dashed line),
and based on Monte–Carlo simulations (diamonds).
(a) (b)
Fig. 5: Competing pathogens dynamics. (a): Fractions of nodes in I, J, S: origi-
nal AME (solid line); lumped AME (dashed line); Monte-Carlo simulations (di-
amonds). (b): Comparison of pair approximation with Monte-Carlo simulation
We assume (λ1, λ2, λ3) = (6.0,0.5,0.5) with kmax = 60 and γ= 3.0. The initial
distribution is set to (xI(0), xR(0), xS(0)) = (3
5). Again, we find that Monte–
Carlo simulations, original AME, and lumped AME are in excellent agreement
(Fig. 4b). The error curve, however, converges slower to zero than in the SIR
model but it gets fast enough close to it (Fig. 4a). The lumped solution cor-
responds to 1032 clusters with a lumping error of 0.0059 and a runtime of 35
seconds compared to 39711 clusters of the original AME solution the runtime of
which was 1606 seconds.
5.3 Competing Pathogens
We, finally, examine an epidemic model with two competing pathogens [22]. The
pathogens are denoted by I and J and the susceptible state by S:
I S λ2·m[J]
J I λ3
S J λ4
We assume that both pathogens have the same infection rate and differ only in
their respective recovery rates. Specifically, we set
(λ1, λ2, λ3, λ4) = (5.0,5.0,1.5,1.0) and assume network parameters of kmax = 55
and γ= 2.5. The initial distribution is (xI(0), xJ(0), xS(0)) = (0.2,0.1,0.7).
This model is the most challenging case study for our approach. AME solution
and naturally lumped AME are not in perfect alignment with Monte Carlo
simulations (Fig. 5a) and our lumping approach needs a, comparably to the
previous cases, larger number of clusters to get a reasonably good approximation
of the AME (Fig. 6a). The computational gain is, however, large as well. The
lumped solution that comes with an approximation error of 0.02 corresponds to
2135 clusters and a runtime of 961 seconds compared to 30856 clusters and 17974
seconds of the original AME solution. Pair approximation approach (Fig. 5b)
even tough faster (40 seconds) than the lumped AME would have here resulted
to a much larger approximation error than our method (cf. Fig. 5b).
At last, the slow convergence of the error curve makes the competing pathogen
model a good test case the for our stopping heuristic. The heuristic evaluates
the model for three different clusterings (509, 986, 2135 clusters). It stops as
the difference between the two last clusterings is smaller than stop, showing
its effectiveness also for challenging models. In Fig. 6b we show the alignment
between the true lumping error and the surrogate error used by the heuristic.
(a) (b)
Fig. 6: Competing pathogens lumping. (a): Lumping error and runtime of the
ODE solver w.r.t. the number of clusters. (b): Lumping error compared to the
error used by the heuristic.
6 Conclusions and Future Work
In this paper, we present a novel model-reduction technique to overcome the
large computational burden of the multistate AME and make it tractable for
real world problems. We show that it is possible to describe complex global
behavior of dynamical processes using only an extremely small fraction of the
original equations. Our approach exploits the high similarity among the original
equations as well as the comparably small impact of equations belonging to the
tail of the power-law degree distribution. In addition, we propose an approach
for finding a reasonable trade-off between accuracy and runtime of our method.
Our approach is particularly useful in situations where several evaluations of
the AME are necessary such as for the estimation of parameters or for model
For future work, we plan to develop a method for on-the-fly clustering, which
joins equations and breaks them apart during integration. This would allow the
clustering to take into account the concrete (local) dynamics and to analyze
adaptive networks with a variable degree distribution.
7 Acknowledgments
This research was been partially funded by the German Research Council (DFG)
as part of the Collaborative Research Center “Methods and Tools for Under-
standing and Controlling Privacy”. We thank James P. Gleeson for his comments
regarding the performance of AME on specific models and Michael Backenk¨ohler
for his comments on the manuscript.
1. Albert-L´aszl´o Barab´asi. Network science. Cambridge university press, 2016.
2. Albert-L´aszl´o Barab´asi and R´eka Albert. Emergence of scaling in random networks.
science, 286(5439):509–512, 1999.
3. Alain Barrat, Marc Barthelemy, and Alessandro Vespignani. Dynamical processes
on complex networks. Cambridge university press, 2008.
4. Luca Bortolussi, Jane Hillston, Diego Latella, and Mieke Massink. Continuous
approximation of collective system behaviour: A tutorial. Performance Evaluation,
70(5):317–349, 2013.
5. Fred Brauer. Mathematical epidemiology: Past, present, and future. Infectious
Disease Modelling, 2(2):113–127, 2017.
6. Peter Buchholz. Exact and ordinary lumpability in finite markov chains. Journal
of applied probability, 31(1):59–75, 1994.
7. Luca Cardelli, Mirco Tribastone, Max Tschaikowski, and Andrea Vandin. Erode:
a tool for the evaluation and reduction of ordinary differential equations. In Inter-
national Conference on Tools and Algorithms for the Construction and Analysis of
Systems, pages 310–328. Springer, 2017.
8. Luca Cardelli, Mirco Tribastone, Max Tschaikowski, and Andrea Vandin. Syntactic
markovian bisimulation for chemical reaction networks. In Models, Algorithms,
Logics and Tools, 2017.
9. Claudio Castellano and Romualdo Pastor-Satorras. Thresholds for epidemic
spreading in networks. Physical review letters, 105(21):218701, 2010.
10. Eric Cator and Piet Van Mieghem. Second-order mean-field susceptible-infected-
susceptible epidemic threshold. Physical review E, 85(5):056111, 2012.
11. Wesley Cota and Silvio C Ferreira. Optimized gillespie algorithms for the sim-
ulation of markovian epidemic processes on large and heterogeneous networks.
Computer Physics Communications, 219:303–312, 2017.
12. G Demirel, F Vazquez, GA B¨ohme, and T Gross. Moment-closure approximations
for discrete adaptive networks. Physica D: Nonlinear Phenomena, 267:68–80, 2014.
13. Nick Fedewa, Emily Krause, and Alexandra Sisson. Spread of a rumor. Society for
Industrial and Applied Mathematics. Central Michigan University, 25, 2013.
14. P. G. Fennell. Stochastic processes on complex networks: techniques and explo-
rations. PhD thesis, University of Limerick, 2015.
15. Bailey K Fosdick, Daniel B Larremore, Joel Nishimura, and Johan Ugander.
Configuring random graph models with fixed degree sequences. arXiv preprint
arXiv:1608.00607, 2016.
16. James P Gleeson. High-accuracy approximation of binary-state dynamics on net-
works. Physical Review Letters, 107(6):068701, 2011.
17. James P Gleeson. Binary-state dynamics on complex networks: Pair approximation
and beyond. Physical Review X, 3(2):021004, 2013.
18. James P Gleeson, Sergey Melnik, Jonathan A Ward, Mason A Porter, and Pe-
ter J Mucha. Accuracy of mean-field theory for dynamics on real-world networks.
Physical Review E, 85(2):026106, 2012.
19. Istvan Z Kiss, Joel C Miller, and P´eter L Simon. Mathematics of epidemics on
networks: from exact to approximate models. Forthcoming in Springer TAM series,
20. Charalampos Kyriakopoulos, Gerrit Grossmann, Verena Wolf, and Luca Borto-
lussi. Lumping of degree-based mean-field and pair-approximation equations for
multistate contact processes. Physical Review E, 97(1):012301, 2018.
21. Genyuan Li and Herschel Rabitz. A general analysis of approximate lumping in
chemical kinetics. Chemical engineering science, 45(4):977–1002, 1990.
22. Naoki Masuda and Norio Konno. Multi-state epidemic processes on complex net-
works. Journal of Theoretical Biology, 243(1):64–75, 2006.
23. Mark EJ Newman. The structure and function of complex networks. SIAM review,
45(2):167–256, 2003.
24. Romualdo Pastor-Satorras, Claudio Castellano, Piet Van Mieghem, and Alessandro
Vespignani. Epidemic processes in complex networks. Reviews of modern physics,
87(3):925, 2015.
25. Romualdo Pastor-Satorras and Alessandro Vespignani. Epidemic spreading in
scale-free networks. Physical review letters, 86(14):3200, 2001.
26. Mason Porter and James Gleeson. Dynamical systems on networks: A tutorial,
volume 4. Springer, 2016.
27. eter L Simon, Michael Taylor, and Istvan Z Kiss. Exact epidemic models on
graphs using graph-automorphism driven lumping. Journal of mathematical biol-
ogy, 62(4):479–508, 2011.
28. James Wei and James CW Kuo. Lumping analysis in monomolecular reaction sys-
tems. analysis of the exactly lumpable system. Industrial & Engineering chemistry
fundamentals, 8(1):114–123, 1969.
A Simplification of Equation Generation
We constructed the lumped equations such that they can be evaluated efficiently
in each step of the ODE solver. However, for very large kmax, the size of M
becomes enormous. This makes the generation of the equations costly because
we iterate multiple times over M. This iteration is necessary, each time we
compute a scalar of the form PmCwC,kmf(m) . In this section we introduce an
approximative scheme to generate lumped equations without the computational
burden of looking at each individual m∈ M.
Our main idea is to only consider the center of each cluster w.r.t. wC,km, and
not at each cluster element. We use hmiCto denote the center of cluster C, each
entry being defined as:
hmiC[s] = X
We can efficiently compute hmiCby only considering the direction of a cluster
(which only depends on the associated proportionality cluster) and the mean
degree of a cluster (which can be computed by only considering the degree dis-
Next, we approximate the average cluster rate by only evaluating the rate
function of each rule at the cluster mean:
This, of course, only makes sense if the rate function is reasonably smooth (which
is the case in our models).
Likewise, inside the βss1ss2, we approximate:
f(m)wC,kmm[s]fhmiC·wC,km· hmiC[s].(20)
Finally, we approximate the in- and outflow related to the βs. Note that, by
design, our clustering has the property that for given s1, s2∈ S and C∈ C there
is only exactly one neighboring cluster in which probability mass can flow by
adding (resp. subtracting) state s1(resp. s2) from the neighborhood vector. We
now assume a fixed s1,s2and use C0∈ C to denote this cluster. We define
CNB ={m|m{s
We see that all mCNB occur in the third term (inflow) and in the fourth
term (outflow) in the lumped AME. Hence, they cancel out and we can ignore
them. To determine the flow of probability mass between two clusters only CB
is of interest. Since the flow is symmetrical (i.e., the inflow of C0is the outflow
of C) it is sufficient to approximate one direction. We use
L·zs,C0· hmiCB[s1]·|CB|
to approximate the flow from Cto C0. Hence, we add (resp. substract) this
value in the equation corresponding to C0(resp. C). Note that hmiCBdenotes
an approximation of the mean value of CB. As these points lie at the border to
C0, we use:
hmiCB[s1] = 1
2hmiC[s1] + 1
|C|is a scaling factor which corresponds to the size of the border between
the clusters (the larger the border area, the more probability flows between the
clusters). Note that also the cardinality of Cand the cardinality of CBcan
be efficiently approximated by combinatoric reasoning without looking at the
individual elements.
First, consider |CB|. If we fix a k, each cluster has approximately the same
number of elements (cf. Fig. 2b). We get this number by dividing |Mk|(the
number of neighbor vectors for that degree) by the number of proportionality
clusters. To determine |C|, we simply aggregate this value over all degrees which
occur inside C. Next, consider |CB|. It denotes the number of points inside C
but next to one particular neighboring cluster. Luckily, our clustering has a nice
geometrical structure (namely a triangular one), which we exploit here. The size
of a face (surface area in one direction) of each cluster C∈ C in ndimensions
for a fixed degree kis exactly the number of elements in a cluster for n1
dimensions for that k. For example, the face of a tetrahedron is a triangle, and
the size of the triangle can be determined by clustering three dimensions instead
of four.
We present two examples of approximative equation generation in Fig. 7.
First, we compare these with equations which are generated using the old ap-
proach (Fig. 7a). We find that their respective dynamics does not differ signifi-
In addition, we test the approximative equation on a model for which the
traditional generation would introduce a significant overhead and where testing
the original AME is practically impossible (Fig. 7b). We choose a SIR model
(where nodes are trapped in state R) with kmax = 500, γ= 2.5, an infection rate
of 3.0·m[I], and a recovery rate of 0.3. Again, we see that the lumped AME is
in excellent agreement with the numerical simulations.
(a) (b)
Fig. 7: Dynamics of approximative equations. (a): The same SIR model as in
Fig. 3: Lumping of the AME (solid line) and approximation of the lumped equa-
tion (dashed line) corresponding to 10 proportionality clusters and 20 degree
clusters. (b): New SIR model with 50 degree clusters and 15 proportionality
clusters. Approximative equations (solid line) are compared with Monte-Carlo
simulations (diamonds). The lumped AME has 8583 clusters. In contrast to more
than 21 million clusters of the original model.
... This work proposes an approximate lumping scheme. Approximate lumping has been shown to be useful when applied to mean-field approximation approaches of epidemic models like the degree-based mean-field and pair approximation equations [32], as well as the approximate master equation [20,15]. However, mean-field equations are essentially inflexible as they do not take topological properties into account or make unrealistic independence assumptions between neighboring nodes. ...
... Moreover, literature is very rich in proposed moment-closure-based approximation techniques for MPMs, which can now be utilized [49,19]. We also plan to investigate the relationship between lumped mean-field equations [20,32] and coarse-grained counting abstractions further. ...
Full-text available
Stochastic processes on complex networks, where each node is in one of several compartments and neighboring nodes interact with each other, can be used to describe a variety of real-world spreading phenomena. However, computational analysis of such processes is hindered by the enormous size of their underlying state space. In this work, we demonstrate that lumping can be used to reduce any epidemic model to a Markov population model (MPM). Therefore, we propose a novel lumping scheme based on a partitioning of the nodes. By imposing different types of counting abstractions, we obtain coarse-grained Markov models with a natural MPM representation that approximate the original systems. This makes it possible to transfer the rich pool of approximation techniques developed for MPMs to the computational analysis of complex networks' dynamics. We present numerical examples to investigate the relationship between, the accuracy of the MPMs, the size of the lumped state space, and the type of counting abstraction.
Full-text available
Contact processes form a large and highly interesting class of dynamic processes on networks, including epidemic and information spreading. While devising stochastic models of such processes is relatively easy, analyzing them is very challenging from a computational point of view, particularly for large networks appearing in real applications. One strategy to reduce the complexity of their analysis is to rely on approximations, often in terms of a set of differential equations capturing the evolution of a random node, distinguishing nodes with different topological contexts (i.e., different degrees of different neighborhoods), like degree-based mean field (DBMF), approximate master equation (AME), or pair approximation (PA). The number of differential equations so obtained is typically proportional to the maximum degree kmax of the network, which is much smaller than the size of the master equation of the underlying stochastic model, yet numerically solving these equations can still be problematic for large kmax. In this paper, we extend AME and PA, which has been proposed only for the binary state case, to a multi-state setting and provide an aggregation procedure that clusters together nodes having similar degrees, treating those in the same cluster as indistinguishable, thus reducing the number of equations while preserving an accurate description of global observables of interest. We also provide an automatic way to build such equations and to identify a small number of degree clusters that give accurate results. The method is tested on several case studies, where it shows a high level of compression and a reduction of computational time of several orders of magnitude for large networks, with minimal loss in accuracy.
Full-text available
Numerical simulation of continuous-time Markovian processes is an essential and widely applied tool in the investigation of epidemic spreading on complex networks. Due to the high heterogeneity of the connectivity structure through which epidemics is transmitted, efficient and accurate implementations of generic epidemic processes are not trivial and deviations from statistically exact prescriptions can lead to uncontrolled biases. Based on the Gillespie algorithm (GA), in which only steps that change the state are considered, we develop numerical recipes and describe their computer implementations for statistically exact and computationally efficient simulations of generic Markovian epidemic processes aiming at highly heterogeneous and large networks. The central point of the recipes investigated here is to include phantom processes, that do not change the states but do count for time increments. We compare the efficiencies for the susceptible-infected-susceptible, contact process and susceptible-infected-recovered models, that are particular cases of a generic model considered here. We numerically confirm that the simulation outcomes of the optimized algorithms are statistically indistinguishable from the original GA and can be several orders of magnitude more efficient.
Full-text available
We give a brief outline of some of the important aspects of the development of mathematical epidemiology.
Full-text available
We give a tutorial for the study of dynamical systems on networks. We focus especially on "simple" situations that are tractable analytically, because they can be very insightful and provide useful springboards for the study of more complicated scenarios. We briefly motivate why examining dynamical systems on networks is interesting and important, and we then give several fascinating examples and discuss some theoretical results. We also briefly discuss dynamical systems on dynamical (i.e., time-dependent) networks, overview software implementations, and give an outlook on the field.
In chemical reaction networks (CRNs) with stochastic semantics based on continuous-time Markov chains (CTMCs), the typically large populations of species cause combinatorially large state spaces. This makes the analysis very difficult in practice and represents the major bottleneck for the applicability of minimization techniques based, for instance, on lumpability. In this paper we present syntactic Markovian bisimulation (SMB), a notion of bisimulation developed in the Larsen-Skou style of probabilistic bisimulation, defined over the structure of a CRN rather than over its underlying CTMC. SMB identifies a lumpable partition of the CTMC state space a priori, in the sense that it is an equivalence relation over species implying that two CTMC states are lumpable when they are invariant with respect to the total population of species within the same equivalence class. We develop an efficient partition-refinement algorithm which computes the largest SMB of a CRN in polynomial time in the number of species and reactions. We also provide an algorithm for obtaining a quotient network from an SMB that induces the lumped CTMC directly, thus avoiding the generation of the state space of the original CRN altogether. In practice, we show that SMB allows significant reductions in a number of models from the literature. Finally, we study SMB with respect to the deterministic semantics of CRNs based on ordinary differential equations (ODEs), where each equation gives the time-course evolution of the concentration of a species. SMB implies forward CRN bisimulation, a recently developed behavioral notion of equivalence for the ODE semantics, in an analogous sense: it yields a smaller ODE system that keeps track of the sums of the solutions for equivalent species.
Conference Paper
We present ERODE, a multi-platform tool for the solution and exact reduction of systems of ordinary differential equations (ODEs). ERODE supports two recently introduced, complementary, equivalence relations over ODE variables: forward differential equivalence yields a self-consistent aggregate system where each ODE gives the cumulative dynamics of the sum of the original variables in the respective equivalence class. Backward differential equivalence identifies variables that have identical solutions whenever starting from the same initial conditions. As back-end ERODE uses the well-known Z3 SMT solver to compute the largest equivalence that refines a given initial partition of ODE variables. In the special case of ODEs with polynomial derivatives of degree at most two (covering affine systems and elementary chemical reaction networks), it implements a more efficient partition-refinement algorithm in the style of Paige and Tarjan. ERODE comes with a rich development environment based on the Eclipse plug-in framework offering: (i) seamless project management; (ii) a fully-featured text editor; and (iii) importing-exporting capabilities.
Random graph null models have found widespread application in diverse research communities analyzing network datasets. The most popular family of random graph null models, called configuration models, are defined as uniform distributions over a space of graphs with a fixed degree sequence. Commonly, properties of an empirical network are compared to properties of an ensemble of graphs from a configuration model in order to quantify whether empirical network properties are meaningful or whether they are instead a common consequence of the particular degree sequence. In this work we study the subtle but important decisions underlying the specification of a configuration model, and investigate the role these choices play in graph sampling procedures and a suite of applications. We place particular emphasis on the importance of specifying the appropriate graph labeling---stub-labeled or vertex-labeled---under which to consider a null model, a choice that closely connects the study of random graphs to the study of random contingency tables. We show that the choice of graph labeling is inconsequential for studies of simple graphs, but can have a significant impact on analyses of multigraphs or graphs with self-loops. The importance of these choices is demonstrated through a series of three in-depth vignettes, analyzing three different network datasets under many different configuration models and observing substantial differences in study conclusions under different models. We argue that in each case, only one of the possible configuration models is appropriate. While our work focuses on undirected static networks, it aims to guide the study of directed networks, dynamic networks, and all other network contexts that are suitably studied through the lens of random graph null models.
Systems as diverse as genetic networks or the World Wide Web are best described as networks with complex topology. A common property of many large networks is that the vertex connectivities follow a scale-free power-law distribution. This feature was found to be a consequence of two generic mech-anisms: (i) networks expand continuously by the addition of new vertices, and (ii) new vertices attach preferentially to sites that are already well connected. A model based on these two ingredients reproduces the observed stationary scale-free distributions, which indicates that the development of large networks is governed by robust self-organizing phenomena that go beyond the particulars of the individual systems.
A lumping analysis of an approximately lumpable monomolecular reaction system is given. One method of constructing the kinetic rate constant matrix for the lumped system is proposed and a method for establishing the magnitude of the errors that accompany lumping is given. Numerical examples illustrate the lumpability concept.
In recent years the research community has accumulated overwhelming evidence for the emergence of complex and heterogeneous connectivity patterns in a wide range of biological and socio-technical systems. The complex properties of real world networks have a profound impact on the behavior of equilibrium and non-equilibrium phenomena occurring in various systems, and the study of epidemic spreading is central to our understanding of the unfolding of dynamical processes in complex networks. The theoretical analysis of epidemic spreading in heterogeneous networks requires the development of novel analytical frameworks, and it has produced results of conceptual and practical relevance. Here we present a coherent and comprehensive review of the vast research activity concerning epidemic processes, detailing the successful theoretical approaches as well as making their limits and assumptions clear. Physicists, epidemiologists, computer and social scientists share a common interest in studying epidemic spreading and rely on very similar models for the description of the diffusion of pathogens, knowledge, and innovation. For this reason, while we focus on the main results and the paradigmatic models in infectious disease modeling, we also present the major results concerning generalized social contagion processes. Finally we outline the research activity at the forefront in the study of epidemic spreading in co-evolving and time-varying networks.