Bootstrap percolation in living neural networks (Q607750)

For a neural network modeled as a random directed graph, one considers at each time step \(n\), the mean of all final states of nodes until \(n\). This gives a proportion of all fired neurons. One shows that the asymptotic, for \(n \to \infty\) , of the final proportion of the fired neurons is dependent on the distribution of in-degree \(Din\) only. This result follows from a theorem for random graphs that satisfy some very special conditions, proved by the author in the paper.