ISNN Special Session List (to be completed)

• Nature Inspired Methods of High-dimensional Discrete Data Analysis

Organizer:
Dusan Husek, Institute of Computer Science, Academy of Science, Czech Republic
Vaclav Snasel, VSB-Technical University of Ostrava, Czech Republic

Description:
High-dimensional data typically takes the form of extracting correlations between data items, discovering meaningful information in data, clustering data items, and finding efficient representations for clustered data, classification, formal concept analysis, and event association. A common problem encountered in disciplines such as statistical data analysis, data analysis, signal processing, textual data representation, and neural network research, is finding a suitable representation of the data in the lower dimension space.
Since the volume (and dimensionality) of data is typically large, the emphasis of new algorithms must be efficient and scalable to large data sets. Analysis of continuous attribute data generally takes the form of eigenvalue/singular value problems (PCA/rank reduction), clustering, least squares problems, etc. Analysis of discrete data sets, however, generally leads to hard problems, especially when physically interpretable results in discrete spaces are desired.
We therefore invite submission from research communities working on different theoretical and applicative aspects of data mining, especially those that are active in cutting-edge frontier topics. These include, but are not limited to:

• binary factor analysis
• market basked analysis
• analysis of voting
• analysis of capital market
• economic analysis
• clustering of semantic features
• discovering pattern in the data
• feature selection for high-dimensional data
• formal concept analysis (FCA).

we also propose publishing of some extended versions of excellent submissions in special issue of international journal Neural Network World (www.cs.cas.cz/nnw). This journal is already for three years indexed by Thomson (ISI Web of Knowledge).

 

 

• Applications of Neural Networks in Electronic Engineering

Organizer:
Professor Qi-Jun Zhang, (qjz@doe.carleton.ca) Department of Electronics, Carleton University,  Canada
Professor Tom Dhaene, (tom.dhaene@ugent.be) Department of Information Technology, Ghent University, Belgium

Description: Artificial Neural Networks (ANNs) are successfully used for the design and analysis of state-of-the-art electrical and electronic systems and components.

This session will highlight various applications of artificial neural networks in different fields of electrical and electronic engineering, and will bring together leading researchers and practitioners from industry and academia to assess the current state and identify new directions. We welcome all papers describing new and original results in this field, and papers providing state-of-the-art overview of a sub-area in the field.
Topics of interest include, but are not limited to:

• Neural and Fuzzy Systems in circuit simulation
• Computational modelling techniques in Electromagnetics
• Design and optimisation
• Surrogate modeling techniques of passive and active devices
• New applications of numerical techniques in Electrical Engineering

 

 

• Fuzzy Neural Networks for Feedback Control Systems

Organizer:
Wen Yu, CINVESTAV-IPN, Mexico (yuw@ctrl.cinvestav.mx)
Yu Tang, National University of Mexico, Mexico

Description:
Both neural networks (NN) and fuzzy logic systems (FLS) are universal estimators. Resent results show that the fusion procedure of these two different technologies has significant advantages over standard feedback controllers for unknown nonlinear systems. Mostly, a neural network or a fuzzy logic system is used to approximate the nonlinearity of the system to be controlled and a controller is synthesized based on universal function approximators (indirect control), or a control law is directly designed using NN, or FLS based on stability theories. Another approach to feedback control design relies on using fuzzy neural networks to approximately solve various nonlinear controller design equations. 
In addition to the classical feedback control theory, adaptive control and robust control are effective techniques to treat system uncertainties but generally suffers from the disadvantage of being able to achieve asymptotical convergence of the tracking error, also the on-line computation load is usual heavy. In robust control designs, a fixed control law based on a prior information on the uncertainties (usually bounds on these uncertainties) is designed to compensate their effects, and exponential convergence of the tracking error to a (small) ball centered at the origin is obtained.
There is a gap between control system community and computational intelligence (e.g. neural networks and fuzzy systems) society.  The purpose of this session is to bring together fuzzy neural networks and feedback control design techniques.
The sub-topics include but not limited to fuzzy neural networks approaches in the following areas:

• Feedback control using neural networks, fuzzy logic and fuzzy neural networks
• Robust neural (fuzzy) control
• Compensation of nonlinearities with (fuzzy) neural networks
• Identification and observers via (fuzzy) neural networks
• Applications of neural (fuzzy) control

 

 

• Advanced Control with Neural Networks

Organizer:
Zengqi Sun, Tsinghua University, P.R.China (szq-dcs@tsinghua.edu.cn)
Fuchun Sun, Tsinghua University, P.R.China (fcsun@tsinghua.edu.cn)

Description:
One of the most important goals in modern control engineering is to find faster and simpler solutions for highly nonlinear control problems in constrained and uncertain environments. In recent years, advanced control system design for complex nonlinear systems has attracted much attention. Many remarkable results in this area have been obtained. Among them, biologically-inspired techniques such as neural networks(NN) have been made particularly attractive and promising for applications to modeling and control of nonlinear systems, owing to its universal approximation abilities, self-organization, learning and adaptation, fault tolerance, parallel distributed abilities. This learning capability of NNs is used to make the controller learn a certain function, highly nonlinear, representing the direct dynamics, inverse dynamics or any other characteristics of the system.
Currently, there emerge a lot of neural networks architectures for control application, such as feed forward network, recurrent network, fuzzy neural networks, wavelet networks, support vector machine, and so on. Those architectures can be used to design adaptive controller, robust controller, sub-optimal controller, fault-tolerant controller, and even more advanced controller. However, some basic problems, such as stability, convergence, and robustness, have not been fully addressed. There remains a huge research space for the neural networks application in control domain.    
Topics of interest include, but are not limited to:

• survey papers on the state-of-the-art of advanced control with neural networks
• new algorithms addressing the problem of neural networks control.
• comparisons of different neural network control approach
• theoretical analysis of existing neural network control algorithms
• applications of neural networks to advanced control systems, including robotic manipulator, process control, spacecraft and so on.

 

 

• Theories and Applications of Cellular Neural Networks (NEW)

Organizer:
Lequan Min, University of Science and Technology Beijing, P.R.China (minlequan@sina.com)

Description:
Cellular Neural Network (CNN) introduced by Leon O. Chua and L. Yang is a new information processing paradigm. In term of its local connectivity, CNNs have been made of CNN universal chips, whose theoretical computation speed can be at least one thousand times faster than the current digital processors. CNNs have been widely studied for practical applications in image and signal processing, robotic and biological visions. The fundamental properties of a CNN need to be described via so-called local rules. However, the local rules of some CNNs with wonderful functions have not been understood completely. Furthermore, some basic properties for some standard CNNs and generalized CNNs have not been proved mathematically, such as stability, convergence, and robustness. The local coupled property of CNN makes it be suitable to describe particularly biological systems. Recently CNN-based complex systems modelling have got much attention. Therefore there are some uncharted engineering, physical, and mathematical territories to be waiting us to explore and exploit for novel applications of CNN.We would like to invite submissions from research societies working on cutting-edge frontier theoretical and applicative topics.
Topics of interest include, but are not limited to:

• VLSI CNN processor and applications.
• CNN-based complex system modelling
• Robust designs for CNN templates
• CNN-based image processing algorithms
• The local activity theories of CNNs and applications
• Stability and convergence of CNN with non-symmetric templates.

 

 

• Pattern Recognition and Information Processing Using Neural Networks (NEW)

Organizer:
Fuchun Sun, Tsinghua University, P.R.China (fcsun@tsinghua.edu.cn)
Huaping Liu, Tsinghua University, P.R.China (hpliu@tsinghua.edu.cn)

Description:
The neural networks based intelligent computing technologies include a range of techniques such as Artificial Intelligence, Perceptual and Pattern Recognition, Evolutionary and Adaptive Computing, Informatics Processing Theories and Applications, Computational Neuroscience & Bioscience, Soft Computing, Case Based and Constrained Reasoning, Agents, Networking and Computer Supported Co-operative Working, Human Computer Interface Issues, etc.
This special session will bring together researchers and practitioners in the area of theory, design, implementation, and applications of neural networks based pattern recognition and information processing. The corresponding directions include feature selection and dimension reduction for regression, learning algorithms, advances in neural network learning methods, learning random neural networks and stochastic agents, self organization, connectionist cognitive science, cognitive machines, neural dynamics and complex systems, computational neuroscience, neural control, reinforcement learning and robotics applications, robotics, control, planning, as well as bio-inspired neural network on-chip implementation and applications.
The successful applications of neural networks in pattern recognition and intelligent information processing suggest that this direction is likely to play an especially important role in science and engineering.
Topics of interest include, but are not limited to:

• Pattern recognition
• Image and signal processing
• Machine learning and learning control systems
• Chaos theory and time series analysis
• Evolutionary algorithms and genetic programming
• Neural science and neural net systems
• Fuzzy set theory and fuzzy neural networks
• Support vector machines and Kernel methods

 

 

• Visual Cognition Using Neural Networks (NEW)

Organizer:
Bo Zhang, Tsinghua University, P.R.China (dcszb@tsinghua.edu.cn)
Changyin Sun, Southest University, P.R.China (cysun@seu.edu.cn)

Description:
Cognitive science is the interdisciplinary study of mind and intelligence, embracing philosophy, psychology, artificial intelligence, neuroscience, linguistics, and anthropology. Its intellectual origins are in the mid-1950s when researchers in several fields began to develop theories of mind based on complex representations and computational procedures.
One of the greatest mysteries of vision is the remarkable ability of the human brain to understand novel scenes and events rapidly and effortlessly. With the development of biology and artificial intelligent technology, more and more researchers focus that how the brain and other intelligent systems adapt to a changing world. This special session aims at researchers who are interested in computational neuroscience, visual cognition, cognitive neuroscience, artificial neural networks, and artificial intelligence.
Topics of interest include, but are not limited to:

• Visual cognition model and theory
• Neural science and artificial neural networks
• Support vector machines and Kernel methods
• Image processing and computer vision
• Feature extraction and pattern recognition
• Multi-modal and multi-source information interaction and fusion
• Target detection and tracking based on visual cognition
• Cognition based data mining and machine learning

 

 

• Swarm Intelligence based Neural Networks and Applications (SIbNNA) (NEW)

Organizer:
Aladdin Ayesh, De Montfort University, UK (aayesh@dmu.ac.uk)
Cyrille Bertelle, Le Havre Universitie, France, (cyrille.bertelle@gmail.com)
Xiaobing Feng, Shanghai Jiao Tong University, China, (fxb@sjtu.edu.cn)

Description:
Swarm intelligence techniques are increasingly popular techniques for optimization researched extensively. Their natural sciences origins make them both computationally inexpensive and scalable where more interaction rules can be added to govern swarm behavior. These simple behavioral rules of interaction lead to complex emergent properties of the system.
As optimization techniques, it has not been long before they are applied to Neural Network optimization for a variety of applications, e.g. image processing. The nature of swarm intelligence techniques make us expect more of this interaction beyond mere optimization of neural nets.
In this special session we look at the interaction between swarm intelligence techniques and algorithms and neural nets. The applications of this interaction and comparison with other models such as GA-NN or Fuzzy-GA-NN will be of great interest. The applications of such models are wide and include but not limited to, image processing, financial forecasting, and data mining. Reports on the applications and performance will be anticipated.

 

• Neural Networks for Networked Control Systems(NEW)
  

Organizer:
Zeng-Guang Hou, Institute of Automation, Chinese Academy of Sciences ,( hou@compsys.ia.ac.cn).
Long Cheng, Institute of Automation, Chinese Academy of Sciences,(chenglong@compsys.ia.ac.cn).

Description:
Recent years witnessed the rapid development of computer, control, and communication. A major feature of modern control systems is to connect several functional subsystems through the communication network together for certain kinds of group goals (behaviors). It is widely recognized that networked control systems will be a challenging future research direction. Meanwhile, neural networks have emerged as a powerful tool for learning, approximation, optimization and computational intelligence. Numerous neural network based control approaches have been proposed for the process control, manipulator tracking, mobile robot navigation, image processing, and pattern recognition. It can be expected that neural networks will find broader application domains in the networked control system.
This special session welcomes both academic and industrial contributions in all aspects of neural network based networked control systems. The topics include, but are not limited to:

• Mobile sensor networks
• Formation control of aircrafts
• Multi-agent coordination
• Information fusion
• Distributed information processing