Architecture of Management Game for Reinforced Deep Learning

Abstract

This article proposes that bona-fide theory for human resource (HR) management connection to performance should include a scientifically approved architecture with explaining power and game theoretical approach that address management behavior tendencies and workplace problems countertendencies to human performance. Management practices have tendencies to improve workers’ human performance. Workplace problems have tendencies to reduce human performance. Game theory is useful because management practices are situation-sensitive, with causal effect on business performance. Deep reinforcement learning with artificial intelligence provides emerging new possibilities, which may revolutionize organizations’ HR-management. This article presents human capital theories with a game theoretical approach. The stochastic Bayesian game seems to be suitable for describing leaders’ behavior meaning to staff performance and annual profit. Using Bayesian management game, it is possible to simulate the management learning outcome where both well-being and business performance flourish. In this case, the managers (players) succeed in achieving the Nash equilibrium between staff quality of working life and sustainable profitability.