NetEvo: A computational framework for the evolution of dynamical complex networks

TL;DR

NetEvo is a flexible computational framework that simulates and evolves dynamical complex networks by optimizing network topology and parameters to improve specified performance measures.

Contribution

It introduces the concept of a supervisor to integrate network rewiring and parameter tuning for studying complex system evolution.

Findings

Enables simulation of dynamical processes on networks

Provides methods for network topology evolution

Facilitates analysis of complex system design

Abstract

NetEvo is a computational framework designed to help understand the evolution of dynamical complex networks. It provides flexible tools for the simulation of dynamical processes on networks and methods for the evolution of underlying topological structures. The concept of a supervisor is used to bring together both these aspects in a coherent way. It is the job of the supervisor to rewire the network topology and alter model parameters such that a user specified performance measure is minimised. This performance measure can make use of current topological information and simulated dynamical output from the system. Such an abstraction provides a suitable basis in which to study many outstanding questions related to complex system design and evolution.