Health Monitoring through an Attention-Based Agent

In the majority of human-machine interaction applications the human model is considered the ideal prototype for machines. In this paper we present an attention-based model, inspired from the human brain, for constructing artificial agents. In this model, attention-required states are identified through irregular patterns and the agent behavior is adapted accordingly. A health-monitoring scenario using such attention-based agents is presented showing the efficiency of the proposed model.