Fixed link

Growth strategy determines network performance

Ana P. Millán, J.J. Torres, S. Johnson, J. Marro

posted on 08 June 2018

pdf ps (1136 views, 1103 download, 0 comments)

The interplay between structure and function is critical in determining the behavior of several systems. Here we propose an adaptive network model inspired in synaptic pruning that couples activity and topological dynamics. The coupling creates a discontinuous phase transition between an ordered memory phase and a disordered one as a function of the transient density. We prove that the existence of an initial transient period with relatively high density is critical in providing ordered stationary states that can be used to store stable memories. We also show that intermediate values of density are optimal in order to obtain these states with a minimum energy consumption, and that ultimately it is the transient heterogeneity in the network what determines the stationary state. Our results here could explain why the pruning curves observed in actual brain areas present their characteristic temporal profiles and, eventually, anomalies such as autism and schizophrenia associated, respectively, with a deficit or excess of pruning.

Submitted by

Recent comments