Toward Learning Systems That Integrate

1 An understanding of learning { the process by which a learner acquires and reenes a broad range of knowledge and skills { is central to the enterprise of building truly adaptive, exible, robust, and creative intelligent systems. Signiicant theoretical and empirical contributions to the characterization of learning in computational terms have emerged from research in a number of disparate research paradigms. The limitations of individual paradigms and of particular classes of techniques within each paradigm are beginning to be recognized. Converging lines of evidence from multiple sources, both theoretical as well as empirical, suggest that artiicial intelligence systems, in order to be able to deal with complex tasks such as recognizing and describing 3-dimensional objects, or communicating in natural language, must be able to eeectively utilize a range of learning algorithms operating with an adequate repertoire of representational structures. This paper draws on a broad range of research on learning in artiicial intelligence, connectionist networks, statistical and syntactic methods in pattern recognition, and evolutionary models to identify the similarities and diierences among, strengths and limitations of, and promising areas for cross-fertilization between, the diierent paradigms.