Integrated Learning Architectures

Research in systems where learning is integrated to other components like problem solving, vision, or natural language is becoming an important topic for Machine Learning. Situations where learning methods are embedded or integrated into broader systems offers new theoretical challenges to ML and enlarge the potential range of ML applications. In this position paper we propose the research topic of integrated learning architectures as an initial discussion of the role of learning in intelligent systems. We review the current state of the art and characterise several dimensions along which integrated learning architectures may vary. This paper has been prepared as a position paper with the purpose of providing an initial common ground for discussion in the ECML-93 Workshop on Integrated Learning Architectures. The paper has been edited by E Plaza on the basis of the individual contributions of the authors. Over the years, AI has divided itself into a number of research areas: planning, learning, vision, knowledge representation etc. Moreover, a multiplicity of learning methods and systems have been developed in the last decade in Machine Learning. There are currently more and more advocates both in ML and AI in general that invite the research community to think about our current situation and re-think our research strategies keeping in mind the long-range goal of a theoretical comprehension and a computational integration of present and future work. In the ML community, a growing trend exists today towards theoretical and implementational integration of the ML methods already developed. We want to bring together in the Integrated Learning Architectures workshop this growing research lines that integrate different ML methods with each other and ML with problem solving. In this paper, we will initiate the discussion of the role of learning in intelligent systems as a key issue on the research agenda on integrated systems.