Special Issue on Advances in Computational Intelligence, Learning and Their Applications
Goutam Chakraborty and Yong Liu
Algorithms based on Computational Intelligence (CI) and algorithms for learning, attracted a large community of researchers in the last four decades or more. In the beginning, such algorithms were motivated to realize human intelligence in machines. They were to explore novelties in research rather than to implement them for products or services. By the end of the 1990s, CI already had a strong theoretical basis. By then, advanced hardware technologies, cheap chip design, miniaturization and reduced power requirement for hardware, encouraged intelligent algorithm applications to proliferate in real products. This trend is growing, and will continue for long in future.
When this special issue was proposed for publication, we were almost certain that submissions on applications would outnumber those on new algorithms on CI or learning per se – and we were right!
The September 2009 Intelligent System Symposium, FAN2009 – the nineteenth in the series – organized at the University of Aizu in Japan, attracted over 100 submissions. Reviewers were asked to comment on the suitability of individual submissions for consideration for this journal issue. We personally checked all recommended submissions, finally settling on just 12 for publication. Authors were asked to expand on their work. Each paper, submitted for this special issue, was further judged by 3 reviewers. Finally, 3 papers were rejected and 9 accepted.
Of the 9 selected, the first deals with Optimization Algorithms. The remaining 8 are on applications. The second and the third papers are on document clustering and recommendation. The fourth and the fifth cover applications on image recognition – posture estimation of human body, and object segmentation using stereo images. The remaining four papers are applications of CI in different ways. The sixth is on similarity computation using collaborative filtering, the seventh deals with interesting stock-market investor behavior, and the eighth on collecting environmental information from wireless ZigBee sensor networks for use in indoor positioning systems. The final paper is an engineering application of CI, tracker for photovoltaic systems using online learning neural networks. All papers selected are interesting and novel in their application areas and approaches. We expect this issue to encourage young researchers to apply CI to novel problems.
We thank the authors for submitting their work and the reviewers for their selfless service. We also thank Reiko Ohta of Fuji Technology Press for her unflagging aid in making this issue a success.