Methods Transfer and Adaptation from Theoretical Physics

^{Copyright: Renormalization group techniques and Feynman diagrams (Helias, 2020 (CC-BY license))}

Collective behavior that emerges from the interplay of large numbers of individual components has traditionally been investigated in statistical physics. For the mathematical formulation of network models and for their analysis we therefore employ, adapt, extend, and develop such methods. Our recent monograph in Springer Nature's Lecture Notes in Physics summarizes this approach that is based on statistical field theory and the theory of disordered systems. These methods are also suitable to cope with underconstrained models, because parameters can be included on a statistical level; for example, predictions can be made based on the knowledge of the statistics of the connectivity, rather than based on the knowledge of each individual synapse.