Date of Original Version
Abstract or Table of Contents
Abstract: "This paper presents a novel approach for the creation of intelligent machine learning programs. It introduces generic learning tasks as the basic learning processes and describes their representation and integration. The role of knowledge in an intelligent machine learning system is identified to a fine granularity. The new approach allows the integration of multiple learning processes into a system that can select the appropriate process based on its background knowledge and the task. Three ways in which the system improves its learning bias over similar classes of problems are identified. The approach is demonstrated in the domain of bridge design."