Intelligence through interaction: Towards a unified theory for learning

• Main Authors: TAN, Ah-hwee, CARPENTER, Gail A., GROSSBERG, Stephen
• Format: text
• Language: English
• Published: Institutional Knowledge at Singapore Management University 2007
• Subjects: Choice Function; Autonomous Vehicle; Adaptive Resonance Theory; Neural Architecture; Category Node; Databases and Information Systems; Numerical Analysis and Computation
• Online Access: https://ink.library.smu.edu.sg/sis_research/6558; https://ink.library.smu.edu.sg/context/sis_research/article/7561/viewcontent/II_ISNN07_av.pdf
• Institution: Singapore Management University

Machine learning, a cornerstone of intelligent systems, has typically been studied in the context of specific tasks, including clustering (unsupervised learning), classification (supervised learning), and control (reinforcement learning). This paper presents a learning architecture within which a universal adaptation mechanism unifies a rich set of traditionally distinct learning paradigms, including learning by matching, learning by association, learning by instruction, and learning by reinforcement. In accordance with the notion of embodied intelligence, such a learning theory provides a computational account of how an autonomous agent may acquire the knowledge of its environment in a real-time, incremental, and continuous manner. Through a case study on a minefield navigation domain, we illustrate the efficacy of the proposed model, the learning paradigms encompassed, and the various types of knowledge learned.