This document summarizes a presentation on machine learning of epidemic processes in networks. It discusses using machine learning to predict epidemic spreading from network structure. Specifically, it covers using features like degree, clustering, and centrality measures as inputs to algorithms like random forests and neural networks to predict the fraction of infected nodes. The best approach uses a combination of network measures, not a single measure. This allows machine learning to help identify influential spreaders and understand how network structure influences epidemic dynamics.
Identifying Appropriate Test Statistics Involving Population Mean
Machine Learning of Epidemic Processes in Networks
1. Machine Learning of Epidemic
Processes in Networks
Francisco Rodrigues
Institute of Mathematics and Computer Science
University of São Paulo
francisco@icmc.usp.br
Workshop on Modelling of Infectious Diseases Dynamics
2. Outline
1. Complex Networks
2. Epidemic processes in networks
3. The influence of network structure
4. Prediction of epidemic processes
5. Challenges and future research
26. Degree-based mean field: SIS model
v
Keeping only the first order terms:
Multiplying the equation with (k–1)pk/〈k〉 and summing over k
characteristic
time
the fraction of
infected neighbors of
a susceptible node k
Epidemic models in networks
27. Degree-based mean field: SIS model
A global outbreak is possible if τ>0, which
yields the condition for a global outbreak as
Satorras andVespignani, PRL, 2001
Epidemic models in networks
28. A. L. Barabási, Network Science, Cambridge, 2015.
0~
Epidemic models in networks
32. Epidemic spreading with awareness
Ventura da Silva et al., Phys. Rev. E 100, 2019
• The rumour and
disease propagate
with different
velocities.
• At each time step:
π : information
1 − π : disease
33.
34.
35. Epidemic spreading with awareness
• If the rumor propagation is too fast, the outbreak
increases!
Ventura da Silva et al., Phys. Rev. E 100, 2019
36. Epidemic spreading with awareness
Applications
• Infectious diseases with no symptoms
(Sexually transmitted diseases (STD)).
37. Spreading depends on the network structure!
Arruda, Rodrigues and Moreno, Physics Reports, 2018
39. = f( ) + E
Hypothesis:
Yi =f(Xi)+εi
f(x) : ℝd
→ ℝ d: number of features
Structure X Dynamics: Prediction
40. Yi =f(Xi)+εi
• The function f is very complicated due to the
presence of non-trivial patterns of connections,
nonlinear effects and correlations between
variables…
f(x) : ℝd
→ ℝ
Structure X Dynamics: Prediction
Rodrigues et al., https://arxiv.org/abs/1910.00544
42. Data
X Y
k(i), cc(i), B(i), PR(i), kc(i), ec(i)
…
k(j), cc(i), B(i), PR(i), kc(i), ec(i)
…
node i
yi
…
yj
…
43. Dynamical processes
• Epidemic spreading: we defi neYi as the
expected fraction of infected nodes when
the disease starts in i
Regression
Rodrigues et al., https://arxiv.org/abs/1910.00544
44. Machine learning:
• To obtain the function
• Random forests
• Neural Networks
f(x) : ℝd
→ ℝ
Rodrigues et al., https://arxiv.org/abs/1910.00544
64. There is no single measure we can use for the
identification of the most influential spreaders!
What measure is the most suitable to
predict disease spreading?
65. A combination of measures is more suitable for
the identification of the most influential
spreaders.
What measure is the most suitable to
predict disease spreading?
66. Machine learning is a very
useful tool to predict
dynamical processes from
the network structure.
67. Challenges
• How to predict disease propagation
from the networks structure.
• The modeling of temporal interactions
and multilayer organization.
• Host heterogeneities.
• Methods for control: quarantine,
vaccination.
• Identification of influential spreaders.