Neural network models of managerial judgment

An experiment is presented which compares actual managerial decision-making, decisions based on neural network models, and decisions based on linear regression rules. Since neural networks are thought to be superior in the case of more complex decisions, two level of environmental complexity are also introduced into the study. The study makes use of data from subjects making a production-scheduling decision. The subject's behavior is modeled with linear decision rules and several neural network models. The results for the experiment are not included in the paper because analysis was incomplete.<
>