|
Research Institute of Intelligent Computer Systems Ternopil National Economic University |
|
2005, Vol. 4, Issue 3 |
|
Contents and abstracts
EDITORIAL Guest Editors: Vladimir A. Oleshchuk 1) and Peter J. A. Reusch 2)
1) Communication and System Security Group, Agder University College
This special issue of the International Scientific Journal of Computing includes a selection of papers presented at the Third IEEE International Workshop on Intelligent Data Acquisition and Advanced Computing Systems: Technology and Applications (IDAACS), which was held in Sofia, Bulgaria, September 5–7, 2005. The workshop was organized by the Institute of Computer Information Technologies of the Ternopil Academy of National Economy of Ukraine and co-organized by the co-organized by the Technical University of Sofia, Bulgaria. EFFICIENCY ESTIMATION OF PARALLEL ALGORITHM OF ENHANCED HISTORICAL DATA INTEGRATION ON COMPUTATIONAL GRID V. Turchenko 1-2), C. Triki 3), L. Grandinetti 1) and A. Sachenko 2)
1) Center of Excellence of High Performance Computing, University of Calabria,
Via P. Bucci 22B, 87036, Rende (CS), ITALY
The main feature of neural network using for accuracy improvement of physical quantities (for example, temperature, humidity, pressure etc.) measurement by data acquisition systems is insufficient volume of input data for predicting neural network training at an initial exploitation period of sensors. The authors have proposed the technique of data volume increasing for predicting neural network training using integration of historical data method. In this paper we have proposed enhanced integration historical data method with its simulation results on mathematical models of sensor drift using single-layer and multi-layer perceptrons. We also considered a parallelization technique of enhanced integration historical data method in order to decrease its working time. A modified coarse-grain parallel algorithm with dynamic mapping on processors of parallel computing system using neural network training time as mapping criterion is considered. Fulfilled experiments have showed that modified parallel algorithm is more efficient than basic parallel algorithm with dynamic mapping, which does not use any mapping criterion. MULTIPLE NEURAL NETWORK MODELS GENERATOR WITH COMPLEXITY ESTIMATION AND SELF-ORGANIZATION ABILITIES El-Khier Bouyoucef, Abdennasser Chebira, Mariusz Rybnik, Kurosh Madani
Image, Signal and Intelligent Systems Laboratory (LISSI / EA 3956), Senart Institute of Technology,
In this article we present a self-organizing hybrid modular approach that is aimed at reduction of processing task complexity by decomposition of an initially complex problem into a set of simpler sub-problems. This approach hybridizes Artificial Neural Networks based artificial intelligence and complexity estimation loops in order to reach a higher level intelligent processing capabilities. In consequence, our approach mixtures learning, complexity estimation and specialized data processing modules in order to achieve a higher level self-organizing modular intelligent information processing system. Experimental results validating the presented approach are reported and discussed. APPLICATION OF OBJECT ORIENTED NEURAL NETWORK TO CONTROL MOTION OF THE LOAD OF A SEA CRANE Pawel Falat 1), Lucyna Brzozowska 2), Krzysztof Brzozowski 3)
1) University of Bielsko - Biala, ul. Willowa 2, 43-309 Bielsko-Biala, Poland, falat@ath.bielsko.pl
The paper presents object oriented approach to design of neural networks. The second part of the article presents an application of the object oriented neural network to control the load of the sea crane of an A-Frame type. The control algorithm has to stabilize load position and compensate the weaving. The model of the A-Frame dynamics were developed and used to achieve the optimal winch drive functions for various sea conditions. Those functions have been used to teach the network. INTRUSION RECOGNITION USING NEURAL NETWORKS Vladimir Golovko1), Pavel Kochurko2)
1) Brest State Technical University, Moskovskaja str. 267, 224017 Brest, Belarus, gva@bstu.by
Intrusion detection techniques are of great importance for computer network protecting because of increasing the number of remote attack using TCP/IP protocols. There exist a number of intrusion detection systems, which are based on different approaches for anomalous behavior detection. This paper focuses on applying neural networks for attack recognition. It is based on multilayer perceptron. The 1999 KDD Cup data set is used for training and testing neural networks. The results of experiments are discussed in the paper. FACE IDENTIFICATION ALGORITHM BASED ON MESH-DERIVED SYNTHETIC LINEAR DESCRIPTORS R. Kh. Sadykhov 1), V. A. Samokhval 2)
1) A Computer Department of the Belarusian State University of Informatics and Radioelectronics, 6 P.Brovka st.,
This paper presents appearance-based face identification algorithm by means of synthetic linear filters. The objective of our research is to construct facial descriptor in the form of linear filter, which should produce high and low outputs for intra- and inter-class recognition problem correspondingly. This filter can be synthesized from 2,5D sparse mesh derived from a given set of images of a person. As ever the filter is created it is then used as facial descriptor, i.e. serves as personal ID for face identification. EVOLUTIONARY DESIGN OF WIND TURBINE BLADES V. Diaz Casas, R. J. Duro, F. Lopez-Pena Integrated Group for Engineering, University of A Coruna (Spain), vdiaz@udc.es, richard@udc.es, flop@udc.es, www.gii.udc.es An automatic design environment is implemented for the aerodynamic design of wind turbine blades. This tool involves the integration of evolutionary techniques and a simple, fast, and robust aerodynamic simulator which was developed for the prediction of the performance of any turbine blade produced by the evolutionary process. The aerodynamic simulator is based on blade element theory in which a panel method is combined with an integral boundary layer code to calculate the blade airfoils’ characteristics. In order to reduce computations some simplifications have been applied and the results corrected by means of the application of neural network based approximations. Results of the simulations obtained using this technique, of the application of the automatic design procedure and of the operation of the wind turbines thus obtained are presented. EVOLUTIONARY ENVIRONMENT FOR 3D MORPHOLOGICAL DESIGN A. Lamas and R.J. Duro Integrated Group for Engineering Universidade da Coruna C/Mendizabal s/n 15403 Ferrol, (A Coruna), Spain {alamas, richard}@udc.es http://www.gii.udc.es This paper deals with the automation of morphological design. The system proposed here is a part of a complete automatic design system that considers the divergent and convergent stages of the design process through evolutionary procedures. The system provides a way to introduce aesthetics in the decision process. This is a difficult problem within an automatic design system, as aesthetics are subjective, depend on the opinions of humans and humans do not usually agree. The objective here is to be able to obtain the best possible aesthetic solutions for a given set of humans. This is achieved through the introduction of a set of man machine interfaces that allow the system to extract information on their relative opinions without explicitly asking them, and combines them with the engineering information provided by other simulators that participate in the design process. ALTERATION CORRECTION IN LYMPHOCYTE IMAGE FOR MICRO NUCLEUSES DETECTION Domenico Luca Carnm, Domenico Grimaldi, Francesco Lamonaca
Dept. of Electronics, Computer and System Sciences, Univ. of Calabria, 87036 Rende –CS, Italy
The paper presents a method pointed out to detect and to correct the alterations of (i) under-exposure, (ii) overexposure, (iii) out of focus, (iv) Gaussian noise, affecting the images acquired into the flow cytometer. These alterations reduce the image quality and interfere with the correct micro nucleus detection in lymphocyte. The objective of the proposed correction is to make the image able to be correctly processed by the pattern matching algorithm (i) to detect the micronucleus into human lymphocytes, (ii) to minimize the doubtful detections, and (iii) to enhance the confidence that in the rejected images are not including the micro nucleuses. The results of numerical tests confirm the validity of the proposed correction method. FAST DETECTION OF MASSES IN MAMMOGRAMS WITH DIFFICULT CASE EXCLUSION Gabor Takacs 1), Bela Pataki 2)
Department of Measurement and Information Systems,
Budapest University of Technology and Economics,
H-1117 Budapest, Magyar tudosok korutja 2.
Breast cancer is one of the most common forms of cancer among women. Currently mammography is the most efficient method for early detection. A simple and fast mammographic mass detection system and two different methods for difficult case exclusion are presented in this paper. The mass detection system uses a modified version of a known algorithm for small masses and a new algorithm for large masses. The first difficult case filtering method is based on tissue density estimation, the second one on mass candidate count. The system was tested with 600 mammographic cases, each containing 4 images. Case-level performance was measured for malignant mass detection first without and then with difficult case exclusion. PROGRESS IN COMPUTATIONAL INTELLIGENCE TO SUPPORT CCTV SURVEILLANCE SYSTEMS Anthony C Davies 1) , Sergio A Velastin 2)
1) Visiting Professor, Faculty of Computing, Information Systems and Mathematics, Kingston University, Penhryn
Road, Kingston, Surrey, KT1 2EE, England (and Emeritus Professor, King’s College London).
e-mail: tonydavies@ieee.org, http://www.tonydavies.org.uk
The development and capabilities of closed circuit television surveillance systems in association with distributed computing systems are reviewed, and the applications to various aspects of surveillance are described. PRIVACY PRESERVING PATTERN MATCHING ON SEQUENCES OF EVENTS Vladimir A. Oleshchuk
Communication and System Security Group, Agder University College
Grooseveien 36, N-4876 Grimstad, Norway
We propose to use pattern matching on data streams from sensors in order to monitor and detect events of interest. We study a privacy preserving pattern matching problem where patterns are specified as sequences of constraints on input elements. We propose a new privacy preserving pattern matching algorithm over an infinite alphabet A where a pattern P is given as a sequence { pi , pi ,..., pim } 1 2 of predicates pi j defined on A . The algorithm addresses the following problem: given a pattern P and an input sequence t, find privately all positions i in t where P matches t. The privacy preserving in the context of this paper means that sensor measurements will be evaluated as predicates ( ) pi ej privately, that is, sensors will not need to disclose the measurements ( ) ( ) ( j ) n j j xi , x2 ,..., x to the evaluator. ANALYTICALLY MODELING UNRELIABLE PARALLEL PROCESSING SYSTEMS WITH GENERAL TASK TIME DISTRIBUTIONS Pierre M. Fiorini 1), Robert W. Rowan 2)
1) University of Southern Maine, Department of Computer Science, Portland, ME, pfiorini@usm.maine.edu
For many computing systems, failure is rare enough that it can be ignored. In other systems, failure is so common that the recovery procedure can have a significant impact on the performance of the system. In this paper, assuming a computing system is unreliable, we discuss how heavy-tail or power-tail job completion time distributions can appear in an otherwise well-behaved task stream. This is an important consideration since it is known that powertails can lead to unstable systems. We then demonstrate how to obtain performance and dependability measures for a class of computing systems comprised of P unreliable processors and a finite number of tasks, N, given different recovery policies. Finally, we discuss the effects of checkpointing on the job completion time distribution. THE EFFECT OF DATA-REUSE TRANSFORMATIONS ON MULTIMEDIA APPLICATIONS FOR APPLICATION SPECIFIC PROCESSORS N. Vassiliadis, A. Chormoviti, N. Kavvadias, S. Nikolaidis
Section of Electronics and Computers, Department of Physics, Aristotle University of Thessaloniki, 54124
Multimedia applications are characterized by a high number of data transfers and storage operations. Appropriate transformations can be applied at the algorithmic level to improve crucial implementation characteristics. In this paper, the effect of data-reuse transformations on power consumption and performance of multimedia applications, realized on an Application Specific Instruction set Processor (ASIP), is examined. An ASIP for multimedia applications designed based on a complete methodology is used to evaluate this effect. Results prove the efficiency of the ASIP solution and indicate benefits from the use of the data-reuse transformations in terms of energy consumption and performance. Also, preliminary results from the exploitation of instruction buffering technique to reduce the energy consumption of the ASIP are presented. A COMPARISON OF AGGREGATION/BROADCAST METHODS AND MULTICOMPUTER ARCHITECTURES, AND AN EXAMINATION OF THE COMMUNICATION OVERHEAD ON THE IBM PSERIES 655 Linda Markowsky
University of Maine, Orono, ME, USA
First, using a simulator, a detailed comparison of the butterfly and direct aggregation/broadcast methods on both hypercube and fully connected multicomputers, both with and without simulation of congestion, is made. Second, the communication overhead an IBM pSeries 655 is examined. MECHANISMS FOR COORDINATION OF MASTER PLANNING AND LOT SIZING WITHIN A HIERARCHICAL PRODUCTION PLANNING MODEL Rainer Leisten 1), Pascal Reusch 2)
1) University Duisburg-Essen, Leisten@uni-duisburg.de
This paper presents a simulation based analysis of different coordination mechanisms in Hierarchical Production Planning. According to the results of the simulations the effects of Aggregation and Capacity Assignments on the overall costs are discussed. SOM BASED DECISION SUPPORT IN FAILURE MANAGEMENT Miki Sirola, Golan Lampi, Jukka Parviainen
Helsinki University of Technology, Laboratory of Computer and Information Science
Computerized decision support system field covers many methodologies and application areas. In this paper Self-Organizing Map (SOM) and knowledge-based techniques are used in combination to reason problematic situations in failure management. A process model that consists of individual connected process components has been developed. A primary circuit of a boiling water nuclear power plant including two branches has been composed. A failure management scenario is thoroughly analyzed and solved with the SOM based decision support system. The structure and reasoning of the Computerized Decision Support System (CDSS) is also shortly discussed. The process model is demonstrated together with the CDSS and shown to be useful. The tool helps operators decision making with various visualizations, and by giving concrete recommendations for possible control actions or other acts. VOICEXML-APPLICATIONS FOR E-COMMERCE AND E-LEARNING Peter J. A. Reusch 1), Bastian Stoll 2), Daniel Studnik 3), Jörg Swade 3)
1) University of Applied Sciences FH-Dortmund Germany, Peter.Reusch@FH-Dortmund.de
VoiceXML is a language of the W3C to create voice-user interfaces, particularly for the telephone. It uses speech recognition and touchtone (DTMF keypad) for input, and pre-recorded audio and text-to-speech synthesis (TTS) for output. The text-to-speech synthesis feature of advanced VoiceXML tools like WebSphere opens new perspectives for e-commerce and e-learning. We are no longer restricted to pre-recorded audio but can bring any text to the ear of the user – a user that could be visually impaired and needs a voice channel to communicate – or a user who can read but who prefers to listen. VoiceXML-applications have been implemented by the authors to support e-commerce (selection of commodities from catalogues) and user guides for hardware (mobile phones, etc.) and software systems (MS project, etc.). New contributions to e-learning are offered. |