Spreading dynamics on complex networks: a general stochastic approach

Area of Science:

• Network science
• Stochastic processes
• Computational biology

Background:

• Network dynamics are often modeled using complex simulations.
• Comparing diverse epidemic models is challenging due to their complexity.
• Markov stochastic processes offer a probabilistic framework for system dynamics.

Purpose of the Study:

• To present a versatile framework for modeling network dynamics using Markov stochastic processes.
• To infer system states by partially describing them through network motifs.
• To provide a common ground for comparing complex epidemics models.

Main Methods:

• Utilizing Markov stochastic processes to model network dynamics.
• Employing network motifs for partial state description and data inference.
• Applying the framework to susceptible-infectious-susceptible (SIS) and susceptible-infectious-removed (SIR) dynamics.

Main Results:

• Demonstrated accurate results for spreading processes at low computational cost.
• Showcased the framework's adaptability for modeling population dynamics.
• Identified a balance between model realism and underlying assumptions.

Conclusions:

• The proposed Markov process framework offers an efficient and versatile approach to network dynamics.
• This method facilitates the comparison of complex epidemic models.
• Explicitly stated assumptions allow for a clear understanding of model limitations and capabilities.