Agents that Reason and Learn

Abstract

The issues related to designing architectures for agents that need to be able to adapt to changing circumstances during deployment are discussed. This research is being carried out in the context of the Smart Internet technology Cooperative Research Center. The first attempt in this project at an architecture involves integrating BDI agent architectures for the reasoning component and reinforcement learning for the learning component. In the research, the learning system used to approximate the Q-function is ALKEMY, a decision-tree learning system with a foundation in higher-order logic. Show more