Recent advances in intelligent paradigms fusion and their applications

Over the past decades, computerized intelligent paradigms have attracted much attention from researchers. Advances in intelligent techniques and approaches, which include neural networks, knowledgebased systems, fuzzy logic, evolutionary algorithms, agent-based techniques, case based reasoning etc, have resulted in many successful applications of these innovative systems. Indeed, applicability of intelligent paradigms to many diverse domains have been demonstrated, which include pattern recognition, image and signal processing, control and automation, robotics, data mining, biomedicine, knowledge management, finance and banking. Instead of employing a single intelligent paradigm, one recent trend of research and development is on the fusion of different intelligent paradigms to tackle complex problems. The combination of more than one intelligent techniques can be realized in many forms, either by a modular cooperation of two (or more) intelligent techniques, each with its own identity; or by fusing one technique into another to form an integrated system. The main rationale behind intelligent paradigms fusion is to exploit the advantages of each intelligent technique and, at the same time, to avoid its limitations, for solving a particular problem. In this special issue, a number of articles which address the usefulness of different intelligent paradigms and their fusions are presented. The intelligent paradigms investigated include fuzzy models, evolutionary algorithms, multi-agent systems, neural networks, particle swam optimization techniques, and artificial immune systems. Applications of these systems are demonstrated in various fields, ranging from real estate appraisal, e-learning, sensor networks, to biomedical data analysis. A summary of the articles covered in this special issue is as follows. Fuzzy models have emerged as an effective tool for intelligent data analysis, where the data can be both quantitative and qualitative. In Król et al., the TakagiSugeno-Kang (TSK) fuzzy model is optimized using evolutionary algorithms. The resulting model is applied to real estate appraisals. Two approaches are examined. The first focuses on learning the rule base while the second deals with learning the rule base and tuning the membership functions simultaneously. The evolutionary algorithms are based on the Pittsburgh approach with real-valued chromosomes comprising the whole rule base or both the rule base and parameters of the membership functions. Applicability of the system is evaluated using a total of 150 real sales transactions made in one of Polish cities. Development in innovative computer technologies has offered many promises in the field of education, especially in distance learning. However, a crucial issue in educational technology is the provision of instructional environments and conditions that comply with different educational goals and learning abilities. In Serçe et al., an Adaptive Intelligent Learning Sys-