International Journal of Information Technology

Vol. 11 No. 10 2005 (Special Issue)

Guest Editorial

Special Issue on Evolutionary Algorithm & Advanced Learning System

Intelligent computing is a quiet fluid concept which is seen to embrace a wide range of techniques that have been attracting some most active and vibrant researches, including the evolutionary computing, neural networks, fuzzy inference, bio-inspired computing, agents, co-operative computing, and human computer interface issues, etc. These techniques have been successfully applied to most sectors of human activities with many fruitful results.

Intelligent computing is a quiet fluid concept which is seen to embrace a wide range of techniques that have been attracting some most active and vibrant researches, including the evolutionary computing, neural networks, fuzzy inference, bio-inspired computing, agents, co-operative computing, and human computer interface issues, etc. These techniques have been successfully applied to most sectors of human activities with many fruitful results.

With the aim of promoting the research, development and application of advanced intelligent computing techniques by providing a vibrant and effective forum across a variety of disciplines, the International Conference on Intelligent Computing (ICIC’2005) was perhaps one of the most comprehensive conferences held in China, 2005, focusing on the various aspects of advances in intelligent computing. Among various sessions covered in ICIC’2005, evolutionary algorithms and advanced learning systems attracted substantial contributions, covering theoretic and methodological exploration and engineering applications. In this special issue, 15 papers have been selected from the ICIC’05 which are in one way or another concerned in the following three general areas relating to evolutionary computing and advanced learning. Because of the diversity of the selected papers, it is difficult to go through in this editorial how the synthesis is achieved in each contribution and only a brief description is provided below to summarize the work of the papers.

1. Evolutionary algorithms and applications

The special issue first select 6 papers that tackle various issues related to evolutionary algorithms. While acknowledging its limited coverage, this collection of papers on evolutionary algorithms offers some interesting contributions.

Gong, Zhou and Li propose a so called cooperative interactive genetic algorithm by combining cooperative genetic algorithm with user’s preference, aiming to reduce user’s workload in genetic algorithm applications.

Gong, Hao, Shi, Shi and Hu then examine the problems of both premature convergence and user’s workload in interactive genetic algorithm applications. They research the methods of extinguishing species based on taboo values and extinguishing individuals based on non-duplicate principle.

Dang, Liu and Zhang present a robust iterative genetic algorithm whose computational complexity is equivalent to the fastest sorting scheme, and they apply their algorithm to the China -Traveling Salesman Problem.

Galily, Roudsari and Riazithe apply genetic algorithms for the design of a Fuzzy Sliding Mode Controller for Active Queue Management (AQM) in computer networks.

Sun and Zhang apply genetic programming for insect pests forecasting in comparison with linear regression models, and report that the GP model could give more reliable result.

Finally, Watts develop an algorithm which uses evolutionary programming to construct fuzzy membership functions with the potential applications to data mining and knowledge-based decision support systems. The evaluation of this algorithm over two well known benchmark data sets gives some promising results.

2. Neural networks applications

In the last decade, we have witnessed a continual growth of research interests from almost all scientific disciplines on neurobiologically inspired artificial neural networks and the associated models, which play a very important role in contemporary artificial intelligence and machine learning. This special issue has selected 4 papers that offer some interesting contributions to neural control, pattern recognition, and other neural network applications.

Jin, Wang, and Xiao investigate the adaptive neural network control of a class of strict-feedback nonlinear systems with unknown the sign of control coefficients. They develop a systematic procedure based on Nussbaum-type function and RBF neural networks which guarantees global stability of the closed-loop systems.

Perng, Ma, Wu, and Lee perform a robust stability analysis of neural control for vehicle steering systems with perturbed parameters. They linearize the neural controller using the classical describing function, and study the stability problem of the equivalent linearized system using the parameter plane method, based on which the amplitude of limit cycles caused by the neural controller can be easily identified.

Park and Moon propose an effective learning method that enables direct optimization of the neural network classifier discrimination performance. They maximize a partial area under a receiver operating characteristic (ROC) or domain-specific curve, which is usually difficult to achieve with classification accuracy or mean squared error (MSE)-based learning methods. They apply the proposed approach to the credit card fraud detection problem to demonstrate its effectiveness.

Sun and Zhang present a strategy for adaptive online sketchy shape recognition using a modified Support Vector Machine (SVM) incremental learning classifier whose effectiveness is demonstrated by various experimental examples.

3. Information processing and fusion

In this special issue, we also collect another 5 papers relating to intelligent information processing and fusion.

Yang investigates the underwater signal classification. He presents a feature extraction method based on principal component analysis in reconstructed state space (RSS-PCA) for acoustic echo classification, with the emphasis on time domain. The experimental results show that the proposed method is able to enhance the current solution and the fusion of the different features promises better performance.

Wang, Li, Chung, and Xu develop an iterative self-adaptive algorithm to filter impulsive noise for color images. Their experimental results show that their method outperforms conventional multi-channel filters in reducing impulsive noise and retaining edges and corners in color images with significantly reduced computational complexity.

Vuskovic and Du investigate the computation of spectral moments of temporal signals. They come up with a new approach which is based on the analysis of the autocorrelations of the original temporal signal, and makes use of the property that the power spectral density is a discrete-time continuous-frequency function.

Huang and Fang are concerned with the application of Micro Electromechanical Systems (MEMS) inertial sensors for the Guidance, Navigation and Control (GNC) of an autonomous Unmanned Aerial Vehicle (UAV). They design a 250-g Micro-GNC for autonomous UAV by integrating MEMS inertial sensors, Global Positioning System (GPS) receiver, Magnetometer and Barometer for the GNC. The real-time flight tests confirm the effectiveness of their proposal.

Rong, Chen, and Ying propose a multi-agent framework for hydraulic motor fault diagnosis based on three-tier architecture that can support a distributed application. Six types of agents implemented by computers are investigated. In particular, the communications between agents are described. They claim that their method can provide a handy trouble-shooting tool that cuts down the time involved in diagnosing failures in hydraulic components in general, and in motors, in particular.

In closing, we hope that the papers selected into in this special issue will serve as a catalyst for further development of evolutionary computing and advanced learning systems.

Acknowledgement

The guest editors wish to thank all the authors for their contributions to this special issue, all the reviewers for their constructive comments, and the conference general Chair, Prof DS Huang from Chinese Academy of Science for his enormous effort in facilitating the post-conference special issues.

Kang Li, Guest Editor
School of Electronics, Electrical Engineering
& Computer Science
Queen’s University of Belfast
Belfast, BT9 5AH, UK

Wen Yu, Guest Editor
Departamento de Control Autom�tico
CINVESTAV-IPN
Mexico D.F., 07360, Mexico

Shuang-Hua Yang, Guest Editor
Department of Computer Science
Loughborough University
Leicestershire, LE11 3TU, UK,

Sanjay Sharma, Guest Editor
School of Engineering
University of Plymouth
Drake Circus, Plymouth, PL4 8AA, UK

Kang Li


Copyright © Singapore Computer Society 2010-2023