Evolutionary Optimisation of Networks towards Complexity

与Erdös和Renyi引入的随机图相比，网络科学提供了在交互拓扑中引入更多复杂性的系统方法。在哪些生物系统中，更高的复杂性会导致更好的恢复力或其他进化益处？这种观点能否为改进技术和社会过程提供洞察力？该项目采用新颖的方法，利用网络复杂性度量来生成日益复杂的网络集合，随后调查网络复杂性如何影响最终网络动态的复杂性。这种方法的动机是一个共同的隐含假设，即复杂的结构有利于自适应系统，并旨在将有益的复杂网络拓扑转化为特定的应用。预计增加的复杂性可以产生有益的影响（如在自然进化的网络中），或可能阻碍功能（预计出现在物流或生产网络中），其中弹性和适应性的好处是以更复杂结构的组织成本为代价的。因此，该项目包括两个面向应用的动态建模研究，其中期望从更高的复杂性中提高网络性能，但也调查了可能挑战这种范式的普遍性的案例。

Network science provides systematic approaches of introducing more complexity in the interaction topology, compared to random graphs as introduced by Erdös and Renyi. In which biological systems does a higher complexity result in improved resilience or other evolutionary benefits? Can this perspective provide insight for improving processes in technology and society? This project takes the novel approach of utilising network complexity measures to generate ensembles of networks of increasing complexity, followed up by an investigation into how network complexity influences the complexity of the resulting network dynamics. This approach is motivated by a common implicit hypothesis that complex structures are beneficial for adaptive systems, and aims to translate back beneficial complex network topologies into specific applications. It is expected that increased complexity can have a beneficial impact (as in networks evolved in nature), or might impede functionality (expected to appear in logistic or production networks) where benefits of resilience and adaptivity come at organisational costs of a more complex structure. The project therefore includes two application-oriented dynamical modelling studies where increased network performance from higher complexity is expected, but also investigates cases which might challenge the generality of such a paradigm.