1991 ... 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024

Contents of Volume 14 (2004)

5/2004 3-4/2004 2/2004 1/2004

6/2004

[1] Kukal J. (Czech Republic): SOM in Metric Space, 469-488.
The Self Organized Mapping (SOM) is a kind of artificial neural network (ANN) which enables the pattern set self-organization in $\mathbf{R}^{n}$ space with Euclidean metrics. Thus, the traditional SOM consists of two layers: input one with n nodes and output one with H ones. Every output node is characterized by its weight vector $w_{k}\in \mathbf{R}^{n}$ in this case. The absence of pattern coordinates in special cases is a good motivation for self-organization in any metric space $\langle \mathbf{U},d\rangle$ . The learning in metric space is introduced on the cluster analysis problem and basic clustering algorithm is obtained. Relationship to the traditional ISODATA method and NP-completeness is proven. The direct generalization comes to SOM learning in metric space, its algorithm, properties and NP-completeness. The SOM learning is based on objective function and its batch minimization. Three estimates of proposed objective function are included. They will help to study the relationship to Kohonen batch learning, cluster analysis and convex programming task. The Matlab source code for SOM in metric space is available in the appendix. Two numeric examples are oriented to self-organization in metric space of written words and metric space of functions.
[2] Svítek M. (Czech Republic): The Decomposition Theory of LTI Systems, 489-506.
The paper presents new methodology how to decompose the high dimensional LTI (linear time invariant) system with both distinct and repeated eigenvalues of transition matrix into a set of first-order LTI models, which could be combined to achieve the approximation of the original dynamics. As a tool, the Sylvester's theorems are used to design the filter bank and parameters of first order models (transition values). At the end, the practical examples are shown and the next steps of research of decomposition theory are indicated.
[3] Ismail I. A., Nabil T. (Egypt): Using Kullback-Leibler Divergence to Predict on Artificial Neural Network, 507-520.
Several algorithms have been developed for time series forecasting. In this paper, we develop a type of algorithm that makes use of the numerical methods for optimizing on objective function that is the Kullbak-Leibler divergence between the joint probability density function of a time series $x_1, x_2, \dots, x_n$ and the product of their marginal distributions. The Gram-charlier expansion is used for estimating these distributions.

Using the weights that have been obtained by the neural network, and adding to them the kullback-Leibler divergence of these weights, we obtain new weights that are used, to forecast the new value of $x_{n+k}$ .
[4] Kvasnička V. (Czech Republic): Holographic reduced representation in artificial intelligence and cognitive science, 521-532.
Holographic reduced representation is based on a suitable distributive coding of structured information in conceptual vectors, whose elements satisfy normal distribution N(0,1/n). The existing applications of this approach concern various models of associative memory that exploit a simple algebraic operation of the scalar product of distributed representations to measure an overlap between two structured concepts. We have described here a method that uses this representation to model a similarity between different concepts and an inference process based on the rules modus ponens and modus tollens.
[5] Coufal D. (Czech Republic): GUHA method Supported Classification of EEG Data, 533-542.
The research reported in the paper is a part of a large project aiming at designing an automatic device for the micro-sleep events detection. In the paper we are interested in the classification of EEG spectrograms with respect to the level of attention (mentation, relaxation, micro-sleep) of a monitored person (a proband). Data mining techniques are used for developing a classification model. Namely, GUHA method is employed for this purpose. It is a method of exploratory data analysis established on logical and statistical bases that has been continuously developed for last 40 years in the Czech Republic.
[6] Book review, 543-546.
[7] Book review, 547-548.
[8] Book review, 549-550.
[9] Book review, 551-552.
[10] Contents volume 14 (2004), 553-556.
[11] Author's index volume 14 (2004), 557-559.

5/2004

[1] Editorial, 365.
[2] Berka P., Laš V., Svátek V. (Czech Republic): NEST: re-engineering the compositional approach to rule-based inference, 367-380.
A compositional approach to rule-based inference is now often considered as overtaken by other approaches. We suggest that a few relatively straightforward extensions together with state-of-the-art implementation techniques should upgrade it to a level making it a useful part of today's knowledge engineering inventory. The ideas developed by the authors in mid-90s have recently been incorporated into a new expert system called NEST. In addition to the traditional network of propositions and compositional rules, NEST also supports binary, nominal and numeric attributes used for derivation of proposition weights, logical (no uncertainty) and default (no antecedent) rules, context expressions and integrity constraints. The inference mechanism combines backward and forward chaining. Uncertainty processing (based on H\'{a}jek's algebraic theory) allows interval weights interpreted as a union of hypothetical cases, and a novel set of combination functions inspired by neural networks has been added. The system is implemented in two versions: stand-alone and web-based client-server one. A user-friendly editor covering all mentioned features is included.
[3] Blaťák J., Popelínský L. (Czech Republic): Mining first-order maximal frequent patterns, 381-390.
The frequent patterns discovery is one of the most important data mining tasks. We introduce RAP, the first system for finding first-order maximal frequent patterns. We describe search strategies and methods of pruning the search space. RAP which generates long patterns much faster than other systems has been used for feature construction for propositional as well as multi-relational data. We prove that a partial search for maximal frequent patterns as new features is competitive with other approaches and results in classification accuracy increase.
[4] Horváth T., Krajči S. (Slovakia): Integration of two fuzzy data mining methods, 391-402.
The cluster analysis and the formal concept analysis are both used to identify significiant groups of similar objects. Rice & Siff's algorithm for the clustering joins these two methods in the case where the values of an object-attribute model are 1 or 0 and often reduce an amount of concepts. We use a certain type of fuzzification of a concept lattice for generalization of this clustering algorithm in the fuzzy case. For the purpose of finding dependencies between the objects in the clusters we use our method of the induction of generalized annotated programs based on multiple using of the crisp inductive logic programming. Since our model contains fuzzy data, it should have work with a fuzzy background knowledge and a fuzzy set of examples - which are not divided clearly into positive and negative classes, but there is a monotone hierarchy (degree, preference) of more or less positive / negative examples. We have made experiments on data describing business competitiveness of Slovak companies.
[5] Ivánek J. (Czech Republic): Using fuzzy logic operators for construction of data mining quantifiers, 403-410.
Relations between two Boolean attributes derived from data can be quantified by truth functions defined on four-fold tables corresponding to pairs of the attributes. Several classes of such quantifiers (implicational, double implicational, equivalence ones) with truth values in the unit interval were investigated in the frame of the theory of data mining methods. In the fuzzy logic theory, there are well-defined classes of fuzzy operators, namely t-norms representing various types of evaluations of fuzzy conjunction (and t-conorms representing fuzzy disjunction), and operators of fuzzy implications.

In the contribution, several types of constructions of quantifiers using fuzzy operators are described. Definitions and theorems presented by the author in previous contributions to WUPES workshops are summarized and illustrated by examples of well-known quantifiers and operators.
[6] Lín V., Dolejší P., Rauch J., Šimůnek Milan (Czech Republic): The KL-Miner Procedure for Datamining, 411-420.
KL-Miner \cite{RSL:04} is a datamining procedure that, given input data matrix $\cal M$ and a set of parameters, generates patterns of the form R ~ C γ. Here R and C are categorial attributes corresponding to the columns of $\cal M$, and γ is a Boolean condition defined in terms of the remaining colums of $\cal M$. The pattern R ~ C γ. means that R and C are strongly correlated on the submatrix of $\cal{M}$ formed by all the rows of $\cal M$ that satisfy γ. What is meant by ''strong correlation'' and how are R, C and γ generated is determined by the input parameters of the procedure. KL-Miner conforms to the GUHA principle formulated in \cite{Ha:78}. It revives two older GUHA procedures described in \cite{Ha:83}: it is very much related to CORREL and contains a new implementation of COLLAPS as a module. In this paper, we mention the motivation that leads to designing of KL-Miner, describing our new implementation of COLLAPS and giving application examples that illustrate the main features of KL-Miner.
[7] Nováková L. Kléma J., Štěpánková O. (Czech Republic): Anachronistic Attributes in Temporal Data: A Case Study, 421-434.
The paper concerns mining data lacking the uniform structure. The data are collected from a number of objects during repeated measurements, all of which are tagged by a corresponding time. No attribute-valued machine learning algorithm can be applied directly on such data since the number of measurements is not fixed but it varies. The available data have to be transformed and preprocessed in such a way that a uniform type of information is obtained about all the considered objects. This can be achieved, e.g., by aggregation. But this process can introduce anachronistic variables, i.e., variables containing information which cannot be available at the moment when a prediction is needed. The paper suggests and tests a method how to preprocess the considered type of data without falling into a trap of introducing anachronistic attributes. The method is illustrated on a case study based on STULONG data.
[8] Takac A. (Slovakia): Cellular Genetic Programming Algorithm Applied to Classification Task, 435-452.
The focus of this paper is the application of the genetic programming framework in the problem of knowledge discovery in databases, more precisely in the task of classification. Genetic programming possesses certain advantages that make it suitable for application in data mining, such as robustness of the algorithm or its convenient structure for rule generation to name a few. This study concentrates on one type of parallel genetic algorithms - cellular (diffusion) model. Emphasis is placed on the improvement of efficiency and scalability of the data mining algorithm, which could be achieved by integrating the algorithm with databases and employing a cellular framework. The cellular model of genetic programming that exploits SQL queries is implemented and applied to the classification task. The results achieved are presented and compared with other machine learning algorithms.
[9] Vomlel J. (Czech Republic): Probabilistic reasoning with uncertain evidence, 453-466.
Bayesian networks became a popular framework for reasoning with uncertainty. Efficient methods have been developed for probabilistic reasoning with new evidence. However, when new evidence is uncertain or imprecise, different methods have been proposed. The original contribution of this paper are guidelines for the treatment of different types of uncertain evidence, the rules for combining evidence from different sources, and the model revision with uncertain evidence.
[10] Book review, 467-468.

3-4/2004

[1] Editorial, 205.
[2] Burnod Y., Duffose M., Frolov A. A., Kadjian A., Řízek S. (France, Russia, Czech Republic): Cooperation between learning rules in the cerebro-cerebellar neural network, 207-220.
Movement learning results from synaptic plasticities in various sites of the brain. Three sites have been particularly studied: the cortico-cortical synapses in the cerebral cortex, the parallel fiber-Purkinje cell synapses in the cerebellar cortex and the cerebello-thalamo-cortical pathway at the level of the thalamo-cortical synapses. We intended to understand how these three adaptive processes cooperate for optimal performance during the arm reaching movement, and how the cerebellar learning is supervised. A neural network model was developed on the basis of two main prerequisites: the columnar organization of the cerebral cortex and the Marr-Albus-Ito theory of cerebellar learning. The synaptic plasticities observed on these sites were incorporated in the model as differential equations. The analytical resolution of the set of rules showed two main results. First, the adaptive processes taking place in different sites do not interfere but complement each other during the learning of the arm reaching movement. Secondly, any linear combination of the cerebral motor commands may generate olivary signals able to supervise the cerebellar learning process.
[3] Coufal D., Turunen E. (Czech Republic, Finland): Short term prediction of highway travel time using GUHA data mining method, 221-231.
We show that prediction of travel time on a 28-km long highway section based on on-line travel time measurements with video is practicable by a data mining method. We introduce a new prediction model, a result of the GUHA style data mining analysis and the Total Fuzzy Similarity method. Comparing the results with the existing Traficon model, our model improves the travel time class prediction. The results obtained by our method are comparable to the MLP neural network model, too.
[4] Kolda T., Faber J., Svoboda P., Dvořák M. (Czech Republic): A model of artificial neuronal networks designed according the natural neuronal brain structures, 233-246.
The functional structure of our new network is not preset; instead, it comes into existence in a random, stochastic manner.

The anatomical structure of our model consists of two input "neurons", hundreds up to five thousands of hidden-layer "neurons" and one output "neuron".

The proper process is based on iteration, i.e., mathematical operation governed by a set of rules, in which repetition helps to approximate the desired result.

Each iteration begins with data being introduced into the input layer to be processed in accordance with a particular algorithm in the hidden layer; it then continues with the computation of certain as yet very crude configurations of images regulated by a genetic code, and ends up with the selection of 10% of the most accomplished "offspring". The next iteration begins with the application of these new, most successful variants of the results, i.e., descendants in the continued process of image perfection. The ever new variants (descendants) of the genetic algorithm are always generated randomly. The determinist rule then only requires the choice of 10% of all the variants available (in our case 20 optimal variants out of 200).

The stochastic model is marked by a number of characteristics, e.g., the initial conditions are determined by different data dispersion variance, the evolution of the network organisation is controlled by genetic rules of a purely stochastic nature; Gaussian distribution noise proved to be the best "organiser".

Another analogy between artificial networks and neuronal structures lies in the use of time in network algorithms.

For that reason, we gave our networks organisation a kind of temporal development, i.e., rather than being instantaneous; the connection between the artificial elements and neurons consumes certain units of time per one synapse or, better to say, per one contact between the preceding and subsequent neurons.

The latency of neurons, natural and artificial alike, is very important as it enables feedback action.

Our network becomes organised under the effect of considerable noise. Then, however, the amount of noise must subside. However, if the network evolution gets stuck in the local minimum, the amount of noise has to be increased again. While this will make the network organisation waver, it will also increase the likelihood that the crisis in the local minimum will abate and improve substantially the state of the network in its self-organisation.

Our system allows for constant state-of-the-network reading by means of establishing the network energy level, i.e., basically ascertaining progression of the network's rate of success in self-organisation. This is the principal parameter for the detection of any jam in the local minimum. It is a piece of input information for the formator algorithm which regulates the level of noise in the system.
[5] Spišiak M., Kozák Š. (Slovakia): Nonlinear predictive control based on artificial neural networks, 247-260.
This paper deals with neural-predictive algorithm for some nonlinear processes in the industry. Neural model predictive control (NMPC) uses artificial neural networks (ANN) for modeling the process and for configuration of the optimizer. The optimizer sets up on-line controller parameters by predicting next control action signals. Depending on the number of prediction steps, the optimizer can predict the process behavior in the future. Therefore this type of predictive control is very useful for the control of the highly nonlinear processes, which are known for their various behaviors. One practical example is the isothermal polymerization reactor where the NMPC controls the output variable very robustly. Finally, this control method is compared with the linear PID controller designed to solve this problem using a genetic algorithm.
[6] Iskandarani M. Z., Shilbayeh N. F. (Amman-Jordan): Classification of herbs using an electronic nose system, 261-273.
An electronic nose system for herbs classification is designed and tested. The system uses the Figaro TGS800 series sensors with an integrated heating element. The testing of the system was carried out using different types of herbs where it was proved to be successful in classifying them into different classes [10, 11]. Database-based software was designed to interface the built hardware and to process the electronic nose signals before being classified.
[7] Brandejsky T. (Czech Republic): Can be creativity reliable?, 275-283.
Presented work after brief discussion of creative reasoning modelling significance for transportation reliability modelling the paper continues by discussion of known applicable techniques of creativity modelling. Because the most significant seems to be analogical and associative reasoning, the unified model of analogical and associative reasoning is presented. The model due to its real-time capabilities enables to model reasoning under condition of processing capacity limitation (and concluding increase of mistaking reactions producing).
[8] Faber J. (Czech Republic): Detection of different levels of vigilance by EEG pseudo spectra, 285-290.
Impaired wakefulness in machine operators poses a danger not only to themselves but often also to the public at large. While on duty, such persons are expected to be continuously, i.e., without interruption, on the alert. For that purpose, we designed and carried out an experimental model of continuous vigilance monitoring using electroencephalography (EEG) and reaction time measured as the latency of the proband's reaction to sound. If constructed, the set together with other logical elements and an alarm can make for an automatic detection of vigilance and, possibly, also of arousal stimuli in cases of microsleep. We have found the following new facts and confirmed the validity of some of the earlier ones.
[9] Novák M. (Czech Republic): Problems of attention decreases of human system operators, 291-301.
The requirements on transportation systems concern not only the quantitative and qualitative aspects of transportation activities, but also still more aspects of their reliability and safety. This concerns not only the transported subjects or goods, but also environment.

The losses caused by failures of transportation activities reach even now a very high level and, if they are not limited by systematic research and preventive activity, they will reach a quite tremendous level soon.

However, practically all the contemporary transportation vehicles, trains, ships and planes and also all the transportation systems need, for their proper operation, interaction with human beings who drive them, control them or use them and maintain them.

In spite of the fact that a significant progress was made in recent years as concerns the transportation systems automation; the fully automatic transportation system in use is still forseen in considerably far future.

Analyzing the reliability and safety of transportation, one finds that the activity of human being is the weakest point. The technical reliability of almost all transportation tools has improved quite a lot in recent years; however, the human subject interacting with them has not changed too much, as concerns his/her reliability and safety of the respective necessary interaction.

Therefore there is an urgent necessity to improve it, and possibilities how to increase it will stay more and more in the focus of our interest.

In this contribution, the overview of the related problems is being made and open problems for further research in this area are discussed.
[10] Svítek M., Pelikán E. (Czech Republic): Monitoring and control of dangerous goods transport, 303-312.
The paper presents the result of the national ITS project "Monitoring and control of dangerous goods transport with help of GNSS (Global Navigation Satellite System)" within which the practical pilot trial on different traffic infrastructure is tested. The presented solution relates to route selection of the dangerous goods transport, so monitoring and control of real movement on selected route is automatically reported.
[11] Svoboda P. (Czech Republic): EEG Analysis and Classification of Microsleep on the Car Simulator, 313-324.
The article addresses the overwhelming problematics of attention decrease of human operators. It presents two sets of experiments used for vigilance detection and possible microsleep prediction. In the first experiment, the analysis of the electroencephalographic activity (EEG) of a human operator (proband) is correlated with the reaction time (RT) to the sound stimulus. For the second set of experiments, the cooperation of a car simulator realized in virtual reality (VR) environment and measurement of EEG is presented. The paper introduces two main methods of the analysis of EEG: frequency analysis and nonlinear analysis based on computation of the state-space trajectory. For the frequency analysis, the delta, theta, alpha and beta bands are computed and compared with nonlinear measures Largest Lyapunov Exponent (LLE) and Correlation dimension (CD). Finally, both experiments are compared and its outcomes are discussed.
[12] Tatarinov V. (Czech Republic): Vigilance classification based on EEG analysis, 325-335.
Decrease of attention and an possible micro-sleep of an artificial system operator is very dangerous and its early detection can prevent great losses. This article deals with a classification of states of vigilance based on the analysis of an electroencefalographic activity of the brain. Preprocessing of data is done by the Discrete Fourier Transform. For the recognition radial basis functions (RBF), a k-nearest neighbor and a method based on the Bayesian theory is used. Its coefficients are found using the maximum likelihood estimation. An experiment with recognition of 6 states of vigilance created according to reaction time is performed.
[13] Votruba Z., Novák M., Veselý J. (Czech Republic): Reliability of interfaces in complex systems, 337-352.
There is common, rather empirically supported knowledge within the body of the System Analysis that complex interfaces (for example "man-machine" interface within the hybrid system, or synapse in the human brain) susceptibly react both on the dimension of the task (i.e.: the number / type / domain of interface parameters / markers), and the level of uncertainty.

In order to quantitatively evaluate this effect, the overview of the different concepts of interface is done first. Then the problem is analyzed on the background of geometrical considerations.

The results of the study indicate that even a low degree of uncertainty has significantly adverse effect on the interface regularity (consequently the reliability of systems processes, as well) if the dimension of the pertinent task is sufficiently high.

Practical implication of this result for system analytics is straightforward - keeping the dimension of the task as low as possible. The interface dimension higher than 5 is in the majority of tasks with moderate uncertainty considerably unfavorable. This result imposes serious constrain to the systems identification.
[14] Vysoký P. (Czech Republic): Changes of driver dynamics caused by fatigue, 353-362.
The main purpose of this contribution is to find indicators of the drivers' fatigue based on compensatory movements of the steering wheel. We focus our attention on changes in delays in drivers' information processing. We examine drivers fatigued by sleep deprivation. To avoid additional disturbances due to complicated dynamics of car steering, the results of simple one-dimensional tracking was analyzed. Preliminary results support using three variables applicable as fatigue indicators.
[15] Book review, 363.

2/2004

[1] Dündar P., Aytac A. (Turkey): Stability of two-dimensional mesh and torus graph, 119-126.
The vulnerability value of a communication network shows the resistance of the network after the disruption of some centres or connection lines until the communication breakdown. In a network, as the number of centres belonging to sub networks changes, the vulnerability of the network also changes and requires greater degrees of stability or less vulnerability. If the communication network is modelled by a graph G, deterministic measures tend to provide a worst-case analysis of some aspects of the overall disconnection process. Differently from other measures, in the neighbour-integrity is considered that any failure vertex affects its neighbour vertices. Neighbour-integrity is very important measure in stability of security networks and spy networks. It replies three questions: How many vertices can still communicate? How difficult is it to reconnect the graph? How can we design an optimal network?

In this paper we discuss the concept of neighbour-integrity. Firstly, we give some definitions and notation and then we calculate some stability numbers of two-dimensional mesh and torus graphs, which are used in computer sciences.
[2] Loo C. K., Rao M. V. C., Rajeswari M. (Malaysia): Growing multi-experts network with confidence interval estimation, 127-137.
A simple and novel method is proposed to estimate the confidence interval of any neural network. A recently introduced Growing Multi-Experts Network (GMN) is embedded with confidence interval estimator whose output directly indicates the defined measure. One-step hybrid learning is employed in which the unsupervised learning method of Growing Neural Gas (GNG) and the supervised learning are implemented simultaneously. Illustrative examples together with the application examples clearly place the utility of the defined measure in sharper focus.
[3] Frolov A. A., Sirota A. M., Húsek D., Muraviev I. P., Polyakov P. A. (Russia, Czech Republic): Binary factorization in Hopfield-like neural networks: Single-step approximation and computer simulations, 139-152.
The unsupervised learning of feature extraction in high-dimesional patterns is a central problem for the neural network approach. Feature extraction is a procedure which maps original patterns into the feature (or factor) space of reduced dimension. In this paper we demonstrate that Hebbian learning in Hopfield-like neural network is a natural procedure for unsupervised learning of feature extraction. Due to this learning, factors become the attractors of network dynamics, hence they can be revealed by the random search. The neurodynamics is analysed by Single-Step approximation which is known \cite{FROHUM97} to be rather accurate for sparsely encoded Hopfield-network. Thus, the analysis is restricted by the case of sparsely encoded factors. The accuracy of Single-Step approximation is confirmed by computer simulations.
[4] Fiori S., Rossi R. (Italy): Statistical characterization of some electrical and mechanical phenomena by a neural probability density function estimation technique, 153-176.
The present paper concerns the estimation of probability density functions using the particular parameterized class of distribution functions implemented by a single non-linear neuron, introduced in the previous contribution \cite{fio:ijns}. The estimation procedure is applied to the statistical characterization of some electrical and mechanical phenomena.
[5] Rai C. S., Singh Y. (India): A statistical approach to Blind Source Separation, 177-185.
Different methods for Blind Source Separation (BSS) have been recently proposed. Most of these methods are suitable for separating either a mixture of sub-Gaussian source or a mixture of super-Gaussian sources. In this paper, a unified statistical approach for separating the mixture of sub-Gaussian and super-Gaussian source is proposed. Source separation techniques use an objective function to be optimized. The optimization process requires probability density function to be expressed in the terms of the random variable. Two different density models have been used for representing sub-Gaussian and super-Gaussian sources. Optimization of the objective function yields different nonlinear functions. Kurtosis has been used as measure of Gaussianity of a source. Depending upon the sign of kurtosis one of the nonlinearities is used in the proposed algorithm. Simulations with artificially generated as well as audio signals demonstrate effectiveness of the proposed approach.
[6] Lim W. S., Rao M. V. C., Loo C. K. (Malaysia): Sequential learning neural network for sonar target differentiation, 187-197.
In this paper, processing of sonar signals has been carried out using the Minimal Resource Allocation Network (MRAN) and the Probabilistic Neural Network (PNN) in differentiating of commonly encountered features in indoor environments. The stability-plasticity behavior of both networks has been investigated. The experimental result shows that the MRAN possesses lower network complexity but experiences higher plasticity in comparison with PNN. The study also shows that the MRAN performance is superior in terms of on-line learning to PNN.
[7] Book review, 199-200.
[8] Book review, 201-203.

1/2004

[1] Editorial, 1-2.
[2] Bouchner P. (Czech Republic): Driver's micro-sleeps detection using virtual simulation, 3-15.
This paper introduces the cooperation of a virtual car simulator and EEG measurements to test a human driver's behavior in demanding situations. After a short explanation of the main principles and tasks of EEG measurements, basic concepts of our experiments are presented. The following part is devoted to problems and solutions concerning the physical model, graphical and other aspects of our simulator. At the end of the article various measuring procedures are presented.
[3] Honců M. (Czech Republic): The cost of sleep-related road traffic accidents in the Czech Republic, 17-19.
The road traffic accidents (RTA) cause large damage on human health and life, material and environmental damages. The human resource losses represent the main component of the social costs of RTA. These total costs are estimated at per cents of GDP. The severity and effects of the sleep-related RTA are similar to the alcohol-related RTA. According to foreign studies they comprise 1 to 25% of all accidents. In the Czech Republic these data are not available, the amount of social costs of the sleep-related RTA can be estimated at billions of CZK yearly.
[4] Hrnčíř Z., Komárek V. (Czech Republic): Analyses of EEG recordings, 21-25.
The article introduces some activities on Child Neurology Clinic in evaluation and utilisation of analyses of EEG recordings concerning diagnostic support and research works in epileptology, autism and other indications. This activity is also a base of our contribution in the project ME701 "Neuroinformatics Bases Creation and Knowledge Data Mining". A short description of the main approaches, utilised sources, used methods and technical means is presented.
[5] Hrubeš P. (Czech Republic): Car simulator scene based on real world geographical data, 27-36.
This paper presents the idea of using two-dimensional geographical data as a base for building a more realistic three-dimensional virtual scene. The background of the geographical information system (GIS) and its stay of art is discussed here. Then a framework architecture with description of system blocks and their functions to realize two to three dimensional transformation is defined here.
[6] Novák M., Votruba Z. (Czech Republic): Challenge of human factor influence for car safety, 37-47.
Human society needs still more intensive exploitation of all kinds of transportation facilities. This need has already lasted for several decades and will be much more imperative in future. Mobility is one of the basic requirements for survival, besides the energy and food resources, health care and security.

The requirements on transportation systems concern not only the quantitative and qualitative aspects of transportation activities, but still more also the aspects of their reliability and safety. This concerns not only the transported subjects or goods but also the environment.

The losses caused by failures of transportation activities reach even now a very high level and if not limited by systematic research and preventive activity, they will reach quite a tremendous level soon.

However, practically all the contemporary transportation vehicles, trains, ships and planes and also all the transportation systems need, for their proper operation, the interaction with human beings, which drive them, controls them or uses them and maintains them.

In spite of the fact that significant progress was made in recent years as concerns the transportation systems automation, the fully automatic transportation system in use is still foreseen in the considerably far future.

Analyzing the reliability and safety of transportation, one finds that the activity of a human being is the weakest point. The technical reliability of almost all transportation tools has improved quite a lot in the past years; however, the human subject interacting with them has not changed too much, as for his/her reliability and safety of the respective necessary interaction.

Therefore there is a vital necessity to improve it and the possibilities how to implement it will stay more and more in the focus of our interest.

In this paper an overview of the related problems is made, the challenges for further research and development in this area are discussed and the outline of a vision, with respect to human interaction reliability, of optimized transportation systems is presented.
[7] Novák M., Votruba Z., Faber J. (Czech Republic): Impacts of driver attention failures on transport reliability and safety and possibilities of its minimizing, 49-65.
Transportation of people and goods represents a still more significant component of the human culture. Its influence is extremely high today and will increase greatly in the future.

Almost all the contemporary transportation systems are based on the necessity of interaction between the transportation tool (vehicle, plane, ship), the transport control system and the human subject. Though a large effort is put into the development of automatic transport systems, none of the present attempts is fully automatic, in all of them the human subject plays a non-neglectable role with considerably high impact on the reliability and safety of the transportation function. Among such functions the driving and control activity dominates.

The drivers, pilots, captains and transportation systems dispatchers and controllers are usually exposed to considerably long and exhausting services, which could last up to 8 and even more hours.

It is generally known that the human subject is not able to retain in the state of vigilance without brakes and relaxation. Usually, its ability to concentrate his/her attention to some activity (like driving of system control) decreases considerably soon, mainly after 45 or 60 minutes only.

The decrease of human subject attention in the course of his/her activity is not monotone of course, it can involve several periods of temporary increases and decreases. However, without exception, if the exposition is long enough, the subject attention finally falls under the limit of acceptability for safe and reliable activity of the particular type. The subject activity becomes dangerous for him/her, his/her environment and for the driven vehicle, plane, ship or the controlled transportation system too.

Finally, the subject falls in the stage of the so-called micro-sleep, in which he/she is not able to produce the particular driving or control activity at all.

A considerably large effort was given to the analysis of negative impacts of this factor. Unfortunately, the used methodology for such analyses differs up to now in many countries, so that the results are not quite comparable. However, one can estimate that between 15 and 40% of all the accidents on the roads are caused by the non-satisfactory level of the human subject attention. If we take into account that the average economic loss of one mortal road accident is estimated to more than 1 million Euro and if the density of such accidents is taken into account as well, we come to a tremendous figure, which has to be enlarged more by the estimation of losses of non-mortal accidents.

This fact is a motivation for considerably intensive research focused on tools for minimizing and prevention of these losses.

In our contribution, we concentrate on the problems of analysis of acceptable human subject attention limits to the possibility of its detection and to prediction of its decrease.

The methodology, on which our research in this area is based, uses the advantage of the sophisticated analysis of human subject electroencephalographic (EEG) signals, which seems to be more suitable than other approaches oriented to such physiological factors, like the eye movement, face analysis, electrical resistance of skin etc., especially because of the higher specification and faster response.

We show that the relevant data of individual nature for each particular person, which remains considerably stable and stored in a special data-base, can be used for determination of the boundaries of the respective individual regions of an acceptable attention level.
[8] Hořínek D., Šonka K., Dostálová S., Pretl M., Faltýnová E. (Czech Republic): Epworth sleepiness scale in sleep apnoea syndrome patients, 67-72.
The Epworth sleepiness scale (ESS) is a short questionnaire designed to quantify subjective sleepiness.

No correlation was found between the ESS values and the selected parameters of the PolyMESAM all-night sleep ventilation test (apnoea/hypopnoea index - number of apnoeas and hypopnoeas per hour, oxygen desaturation index - number of saturation drops per hour, heart rate variation index - number of heart rate changes per hour, Min Sa02 - mean oxygen saturation minima in percents) in a group of 41 men and 13 women (mean age 48.5 ąSD=9.2) with the sleep apnoea syndrome (SAS).

The mean ESS value in patients with straightforward SAS was 11.1 (ą6.1) while in the control group of 23 men and 6 women (middle age 47.3 ą6.8 years) it was 6.5 (ą2.2). There is a statistically significant difference between the two (p<0.01). In the authors' view, ESS is a useful instrument for testing subjective sleepiness in SAS patients.
[9] Volná J., Kemlink D., Pretl M., Vaňková J., Švejdová K., Šonka K. (Czech Republic): Excessive daytime sleepiness in the sleep apnoea syndrome (SAS) and in a combination of SAS and periodic limb movements in sleep measured by the Multiple Sleep Latency Test, 73-86.
Excessive daytime sleepiness (EDS) is, in its consequences, a major problem for the patient and for the society. EDS is mainly caused by night-time sleep disorders, in particular: the sleep apnoea syndrome (SAS) and periodic limb movements in sleep (PLMS). Our study was designed for finding out a) if there is a correlation between the gravity of the two conditions and the degree of EDS in patients with SAS and those with SAS and PLMS combined, and b) if EDS in the SAS+PLMS group is greater than in patients with the SAS alone. 35 patients with SAS and 10 with SAS+PLMS were examined using nocturnal polysomnography (PSG) and the multiple sleep latency test (MSLT).

As for SAS, no correlation was found between EDS and the gravity of the condition. On the other hand, PLMS was found correlated to the mean sleep latency value. A correlation between the gravity of the disease and the reduced TST, SE and the sleep stages under study was corroborated for both groups. In addition, both conditions were found to interfere with the sleep architecture more than the SAS alone. This noctural sleep disturbance, though, would stop short of raising the mean sleep latency tested in the MLST. Correlations were found in SAS between the age and 2NREM latency and between the latter and average sleep latency. In the MSLT, the groups showed no difference as to the mean sleep latencies.
[10] Svoboda P., Faber J., Tatarinov V. (Czech Republic): Descriptive measures of EEG signal with respect to hypnagogium, 87-96.
Impaired wakefulness of machine operators presents a danger not only for themselves, but often for the public at large as well. While on duty, such persons are expected to be continuously, i.e. without interruption, on the alert. For that purpose, we designed and carried out an experimental model of continuous vigilance monitoring using electroencephalography (EEG) and reaction time measured as the latency of the volunteers' reaction to a sound stimulus. In this article, we focus on two different approaches of EEG signal analysis. Spectral analysis, which is based on linear stochastic approach, is the first type. On the other hand, there is nonlinear analysis formally called the chaos theory. For both methods, we will show typical markers which represent the state of the vigilance. Both methods will be compared and the outputs will be discussed.
[11] Tichý T. (Czech Republic): Fuzzy and expert approaches to detection of micro-sleep, 97-108.
The goal of this paper is to desigen a system for the detection and evaluation of the decrease of drivers' attention. The paper refers to possible reasons and risks of a phenomenon called micro-sleep. It also describes the measurement methodology. The paper apprise of a new way of detection and evaluation of micro-sleep. The methods of fuzzy system and knowledge system is used. The state of the proband is assessed by classification modules F-COMS and C-COMS which were designed specifically for these systems. The paper also describes the individual frequencies of an EEG spectrum including their mutual dependencies in particular states and individual contributions in a detailed way. The conception of the proposed system for detection and evaluation of the decrease of attention ensues from the uniqueness of an individual, and the inaccuracy of measurement on humans, as well as from the universality of a system construction, capable of detecting not only the inattention of a driver, but also his/her maximum concentration or pathological changes.
[12] Vysoký P. (Czech Republic): Changes in car driver dynamics caused by fatigue, 109-117.
Performances and reliability of a human operator are influenced by fatigue. Influence of fatigue on the human operator - car driver as a controller in the steering control loop is discussed. Demonstrated are possibilities of fatigue identification from small compensatory movements of the steering wheel. Preliminary results of using a fatigue indicator based on the analysis of compensatory movements of the steering wheel are introduced.