Detecting Competitive Behaviors in Conflicts

We developed a method for detecting aggressive agents in egress simulations with a cellular-automata model. There are two types of agents, which are normal agents and aggressive agents. Aggressive agents tend to push out others in conflicts and try to move to their target cell aggressively. We considered all the possible combinations of agent types, labeled them, and computed the joint probabilities of the labels from the conflict data obtained from the egress simulations. The label which achieved the maximum joint probability was regarded as the expected label. The accuracy of our method achieved larger than 95% when a few very aggressive agents exist in a group of normal agents. On the other hand, the accuracy decreases when the aggressiveness of aggressive agents decreases or the fraction of the aggressive agents increases.