The main differentiator of the new generation of autonomous systems that is emerging in the twenty-first century is the adaptivity of their intelligence. They are not simply automatic (usually remote) control devices, not only adaptive control systems in the narrow sense of systems with tunable parameters as in the last decades of the past century, but they are rather systems with a certain level of evolving intelligence. While conventional adaptive techniques (Astroem and Wittenmark, 1989) are suitable to represent objects with slowly changing parameters, they can hardly handle complex (usually, nonlinear, nonstationary) systems with multiple operating modes or abruptly changing characteristics since it takes a long time after every drastic change in the system to update model parameters. The evolving systems paradigm (Angelov, 2002) is based on the concept of evolving (expanding or shrinking) system structure that is capable of adapting to the changes in the environment and internal changes of the system itself that cannot solely be represented by parameter tuning/adjustment.

Evolving intelligent systems (eIS) the concept of which was pioneered recently (Angelov, 2002; Kasabov, 2002; Angelov and Kasabov, 2005; Kasabov and Filev, 2006, Jager, 2006), develop their structure, their functionality, and their internal knowledge representation through autonomous learning from data streams generated by the (possibly unknown) environment and from the system self-monitoring. ...

Get Autonomous Learning Systems: From Data Streams to Knowledge in Real-time now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.