Social network learning: Uncertainty, heterogeneity, and the application in principal–agent relationships

Abstract

This paper employs algebraic transformation to describe complex social network learning (SNL) behaviors under continuous expected payoff. Three distinct algorithms are then introduced that factor in uncertainty and heterogeneity. We find that individuals' strategies tend to converge through SNL. We then construct a framework for studying the convergence process in the principal–agent relationship by applying our SNL algorithms to distinct scenarios. Our results show that network topology plays a significant role in changes in the payoffs and the convergence speed of individuals' strategies. We also evaluate the impacts of uncertainty, heterogeneity, agents' output efficiency and risk aversion, and individual's centrality on the effectiveness of SNL.