| An interaction model on random networks: from social to airports networks | | | Philippe Curty | | | Scimetrica, Bern, Switzerland | | |
arXiv:physics/0509074 [database] | | |
Abstract: Social movements, neurons in the brain or even industrial suppliers are best described by agents evolving on networks with basic interaction rules. In these real systems, the connectivity between agents corresponds to the most efficient state of the system. The new idea is that connectivity adjusts itself because of two opposite tendencies: information percolation, decision making or coordination are better when the network connectivity is small. When agents have many connections, the opinion of a person or the state of a neuron tend to freeze: agents find always a minority among their advisors to support their opinion. A general and new model reproduces these features showing a clear transition between the two tendencies at some critical connectivity. Depending on the noise, the evolution of the system is optimal at a precise critical connectivity since, away from this critical point, the system always ends up in a static phase. When the error tolerance is very small, the optimal connectivity becomes very large, and scale free networks are obtained like the airports network and the internet. This model allows for the first time to have a general theory of evolution of agents on networks where evolving scale free networks are now a special case of a larger class of networks. | | |
Type: preprint | | |  |
| | | |
|
