Data streams pose different challenges as compare to the batch sets of data. Yet, most of the existing approaches of data mining, machine learning (including clustering, classification, pattern recognition), process control (including system identification, modelling and control), signal processing (including image/video processing, speech/audio processing), evolutionary robotics etc. are still addressed in off-line mode and the data are assumed to be available in batch form.

At the same time, on-line and incremental approaches existed for quite some time in all of these areas, but they all usually assumed the model/cluster structure/classifier/controller to be with a pre-fixed structure, usually selected subjectively by experience.

Many modern day advanced industrial applications, such as robotics, autonomous systems, various defence and security applications, advanced chemical and petro-chemical processes, Internet applications etc. generate with a high rate data streams that require on-line, real-time processing with models and systems with structure that can not necessarily be pre-fixed. This new challenge was addressed by the new paradigm of the evolving intelligent systems (eIS) coined less than a decade ago. The newly established concept of evolving intelligent systems (eIS) is a result of the synergy between conventional systems, neural networks and fuzzy systems as structures for information representation and real time methods for machine learning. This emerging area targets non-stationary processes by developing novel on-line learning methods and computationally efficient algorithms for real-time applications. To address the problems of modelling, control, prediction, classification and data processing in a dynamically changing and evolving environment, a system must be able to fully adapt its structure and adjust its parameters, rather than use a pre-trained and a fixed structure. That is, the system must be able to evolve, to self-develop, to self-organize, to self-evaluate and to self-improve. Wireless sensor networks, assisted ambient intelligence, embedded soft computing diagnostics and prognostics algorithms, intelligent agents, smart evolving sensors; autonomous robotic systems etc. are some of the natural implementation areas of eIS as a realistic and practical tool for design of real time intelligent systems.

While achieving certain results in addressing the problem of managing uncertainties in data streams with the use of eIS a range of problems are still open and require concerted efforts of the research community to be addressed. An unexhaustive list follows in a form of the topics of interest for the papers that will form the special session: