An Agent Architecture for On-Line Learning of Procedural and Declarative Knowledge

In order to develop versatile cognitive agents that learn in situated contexts and generalize resulting knowledge to diierent environments , we explore the possiblity of learning both declarative and procedural knowledge in a hybrid connectionist architecture. The architecture is based on the two-level idea proposed earlier by the author. Declarative knowledge is represented symbolically, while procedural knowledge is represented subsymboli-cally. The architecture integrates reactive procedures , rules, learning, and decision-making in a uniied framework, and structures diierent learning components (including Q-learning and rule induction) in a synergistic way to perform on-line and integrated learning.