Neural Network Analysis of Behavioral Agent-Based Service Channel Data

When developing an agent-based model for service channel design, the individual decision-making process of the agents is a vital part of the simulation. Additionally, due to the nature of agent-based models and the communication networks that exist between agents, the micro/macro-dynamics are heavily linked. To better understand this link, we propose the use of integrated neural networks trained in a supervised learning environment. Training these networks on data collected from human based experiments, and implementing these neural networks into the model will capture the irrational behavior not captured by traditional models, while improving on traditional agent-based decision-making processes.