Additive Noise-Induced System Evolution (ANISE)

Abstract

Additive noise has been known for a long time to not change a systems stability. The discovery of stochastic and coherence resonance in nature and their analytical description has started to change this view in the last decades. The detailed studies of stochastic bifurcations in the last decades have also contributed to change the original view on the role of additive noise. The present work attempts to put these pieces of work in a broader context by proposing the research direction ANISE as a perspective in the research field. ANISE may embrace all studies that demonstrates how additive noise tunes a systems evolution beyond just scaling its magnitude. The article provides two perspective directions of research. The first perspective is the generalization of previous studies on the stationary state stability of a stochastic random network model subjected to additive noise. Here the noise induces novel stationary states. A second perspective is the application of subgrid-scale modeling in stochastic random network model. It is illustrated how numerical parameter estimation complements and extends subgrid-scale modeling and render it more powerful.

An early warning indicator trained on stochastic disease-spreading models with different noises

“…The most common noise model is the additive noise model, wherein a perturbed observed compartment is construed as the sum of unaltered deterministic dynamics and an additive noise component [49]. This noise model serves as a means to simulate the inherent stochasticity of natural processes and reproduce their consequential impact on a system [50]. To model additive white noise, we add a white noise term into each equation of the deterministic model, encompassing the continuous intensity of fluctuation associated with the dynamics of individual compartments.…”