Variation of motifs
in complex networks and its impact on the information processing capacity
Abstract. This thesis discusses the impact of network motifs on the information processing capacity within graphs. In directed graphs motifs of size 3 are studied and varied using two presented algorithms. Binary cellular automata are applied on the graphs to analyse the dynamics of the network. The information processing capacity is measured by shannon and word entropy, which are determined after every step of variation and visualized in a scatter diagram. A trajectory emerges, which represents the changes in the information processing capacity.
Studying a direct correlation of network motifs and the information processing capacity is the objective of the thesis. Some motif classes show the ability of establishing greater entropies than others. The motif “feed-forward loop” was analyzed particularly in this context. Moreover it turned out, that the used transition function as well as the presence of cycles in the graphs has a big impact on the entropies.