ABSTRACT
We develop a Double Transition Model (DTM) (Yu, 2013; Yu and Santos, 2016) to capture a Commander’s decision-making process and assess the model through evaluations on synthetic and human Commanders in Naval warfare scenarios plus UAV operators in training simulations. A DTM can be used to derive a dynamic Markov Decision Process (dMDP) in which each state reflects a cognitive state of a decision maker that can be discovered/encountered over time and described by his context and each action represents the tasks and decisions that he makes (e.g., during a battle), while the rewards represent the effect of a decision’s outcomes.
