Regensburg 2019 – wissenschaftliches Programm
SOE 12.5: Poster
Dienstag, 2. April 2019, 16:00–19:00, Poster A
From utility functions to many-particle Hamiltonians (and back). — •Michael Schnabel and Daniel Diermeier — University of Chicago, Chicago, USA
As a simple model of binary opinion formation we consider a population of N interacting agents that can select among two different states. We assume that the individual choices are made based on maximizing some underlying utility function that depends on the current distribution of opinions in the population, but otherwise can have arbitrary shape. The resulting stationary distribution of opinions can be written as a Gibbs-Boltzmann distribution with an effective Hamiltonian that describes the interactions occurring at the microscopic scale. We present a method how to obtain the interaction coefficients and describe their scaling behavior as a function of the population size N. Moreover we derive a criterium to tell for what type of utility function the pairwise approximation will be applicable (or not).