An agent architecture with on-line learning of both procedural and declarative knowledge

Conference · Tue Dec 31 04:00:00 EST 1996

OSTI ID:466431

Sun, R; Peterson, T; Merrill, E ^[1]

---

1. Univ. of Alabama, Tuscaloosa, AL (United States)

In order to develop versatile cognitive agents that learn in situated contexts and generalize resulting knowledge to different environments, we explore the possibility 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 subsymbolically. The architecture integrates reactive procedures, rules, learning, and decision-making in a unified framework, and structures different learning components (including Q-learning and rule induction) in a synergistic way to perform on-line and integrated learning.

• OSTI ID: 466431
• Report Number(s): CONF-9610138--
• Country of Publication: United States
• Language: English

Related Subjects

99 GENERAL AND MISCELLANEOUS
ARTIFICIAL INTELLIGENCE
DECISION MAKING
LEARNING