A simple mean field model for social interactions: dynamics, fluctuations, criticality (Q977199)

In the paper we have a study about the dynamics of a spin-flip model with a mean field interaction in the case of social interactions. Such systems are considered as the complex ones and are described in terms of physics. They have own peculiar features: the same information is potentially available to all individuals (geometrical constraints in the interaction are nor justified) and interactions may be more naturally given in a dynamic framework (in contradiction to the physical systems where they are coded by an energy function, i.e. \textit{Hamiltonian}). However in some cases the interactions of \textit{mean field} type may provide accurate descriptions of real behavior; for example when one models the economic behavior of agents sharing the same information. But if one considers for example such situations where individuals tend to conform their choices such an approach can be inappropriate. A typical feature of dynamic mean field models is the existence of \textit{phase transition}: in the limit of infinitely many individuals, different initial conditions may be attracted to different equilibria, i.e. different degrees of opinions polarizations, better or worse financial state, and so on. The authors of paper study in their work a generalization of previously given model (see the Ref. [10]). In the studied model we have a large number \(N\) of interacting individuals, each individual can be in a state \(\omega_i\) and other individuals can \textit{perceive} this state subject to a random perturbation, the change of a state is done by ''adapting'' to the perceived state of all others. The individuals are not homogeneous -- they are divided into reference groups. The paper consists of six sections. First one is a mentioned above introduction that gives an outline of the considered problem. The second one is a short description of a model with an infinitesimal generator \(L_N\) (Eq. (2)) given for a description of 2\(N\)-spin system evolution as a continues Markov chain. In a third one we have the derivation of the process (2) dynamics in the limit as \(N\rightarrow\infty\) in a fixed time interval \([0,T]\) -- this gives the Eq. (4) which describes the behavior of the system governed by mentioned generator (2) in the infinite volume limit (infinitely many individuals). As a result the authors conclude that for example the polarization of opinions may occur or not depending on the properties of initial conditions. In this section we have also the analysis of fluctuations with conclusion that in considered process the dynamics of fluctuations converges weakly to a Gaussian one. In Section 4 we have the considerations about the critical dynamics of the system, i.e., the long-time behavior of the fluctuations in the threshold case; the limiting critical fluctuation process is no trivial one, it is non-Gaussian. Section 5 gives the conclusions and in the last (sixth) Section we have the proofs of Lemmas and Theorems used in this paper.