Parts | Days | Selection | Search | Downloads | Help

AKSOE: Physik sozio-ökonomischer Systeme

AKSOE 2: Dynamics of Groups and Organizations II

AKSOE 2.4: Talk

Friday, March 4, 2005, 15:30–16:00, TU P-N203

Affecting Network Structures Through Selection Mechanisms — •Adrian Marcelo Seufert¹ and Frank Schweitzer² — ¹TU Berlin, Institut für Theoretische Physik — ²D-MTEC ETH Zürich

We present a modification of a model due to Sanjay Jain and Sandeep Krishna of interacting components on a network structure that evolves in time. The nodes in the network can be identified with species in a common environment, and the interactions between different species are represented by links in a directed graph. Each species can be characterized by its population. The nodes are submitted to a simple linear dynamics that affects their populations. The dynamics is determined by the structure of the network, while at the same time influencing it by determining which nodes will be selected for mutation. Mutating nodes lose all their links and randomly set new links, thus modifying the network. A connected network appears in the model once an autocatalytic set appears by chance and engulfs the whole set of nodes.

We modify the model by introducing a new selection mechanism which makes selection dependent on the relative performance of the nodes. We show that by varying the performance threshold required for survival, we affect the structure of the emerging network, as measured by the size of its core, the clustering coefficient, and the out-degree distribution. Also, we observe a separation into two distinct regimes, a high performance regime and a low performance regime, as a function of the threshold.