This paper introduces a paradigm of connectionist architectures and an approach to building on-line, adaptive intelligent systems. This approach is called evolving connectionist systems (ECOS). EGOS evolve through incremental, on-line learning, both supervised and unsupervised. They can accommodate new input data, including new features, new classes, etc. The ECOS framework is presented and illustrated on a particular type of evolving neural networks - evolving fuzzy neural networks. The task of voice recognition and person identification is used as a case study.
Keywords: Evolving connectionist systems, On-line adaptive learning, Multi-modal information processing, Adaptive fuzzy neural networks, Person voice identification