System crash as dynamics of complex networks

Abstract

Complex systems, from animal herds to human nations, sometimes crash drastically. Although the growth and evolution of systems have been extensively studied, our understanding of how systems crash is still limited. It remains rather puzzling why some systems, appearing to be doomed to fail, manage to survive for a long time whereas some other systems, which seem to be too big or too strong to fail, crash rapidly. In this contribution, we propose a network-based system dynamics model, where individual actions based on the local information accessible in their respective system structures may lead to the "peculiar" dynamics of system crash mentioned above. Extensive simulations are carried out on synthetic and real-life networks, which further reveal the interesting system evolution leading to the final crash. Applications and possible extensions of the proposed model are discussed.complex systems | system crash | pseudo-steady state | cascade behavior

Understanding Popularity, Reputation, and Social Influence in the Twitter Society

“…This contrasts with the initial assumption of resilience models for online social networks (Garcia et al, ): a monotonically decreasing inactivity probability with degree. Our results show that a general negative trend is in place, but also show a concavity pattern that calls for new microdynamic formulations, as suggested in the framework of Garcia et al () and recently explored by Yu et al (). Nevertheless, we must note that the cascading inactivity hypothesis follows from these kind of models, which we find support for when comparing the quality of predictors based on degree and coreness.…”

Long noncoding RNA: noncoding and not coded

“…8 Furthermore, it has been demonstrated that lncRNA regulate genes in cis and their transcription per se , not the actual sequence, is activating the adjacent gene. 9 Considering that there are many such lncRNA/protein coding RNA gene pairs in our genome, 10 it is important to determine the extent and functions of protein coding gene regulation by adjacent lncRNA genes.…”

Consensus ranking for multi-objective interventions in multiplex networks

“…Perturbing a node to influence one layer inadvertently affects the other layers that the node is active in, which can have drastic unintended consequences [5,[16][17][18]. Therefore, it is desirable to identify targets in a multiplex network that maximize our influence in certain layers, yet minimize any unwanted impact on others.…”