Construct a Sequential Decision-Making Model: A Dynamic Bayesian Network Perspective

Abstract

Corporate decision makers always face challenges to decide whether to adopt a new technology with a decision often leading to significant subsequence in the corporations. Observational learning theory applies when a person uses observed behavior from others to infer something about the usefulness of the observed behavior. It is known that observational learning may lead to informational cascades when a person ignores his private signals in decision making process. Walden and Browne propose a simulation procedure to model the influence of observational learning in sequential decision making. The objective of this study is to apply a dynamic Bayesian network (DBN) to model decision makers' sequential decision making and observational learning. We show that this DBN perspective of sequential decision making is easy to understand and flexible enough to consider more scenarios not considered in Walden and Browne.