Contents of Volume 20 (2010)

1/2010 2/2010 3/2010 4/2010 5/2010 6/2010


7/2010 - Special Issue on the 5th International Conference on Hybrid Artificial Intelligence Systems

  • [1] Editorial, 807-809.

  • [2] Irigoyen E., Larrea M., Valera J., Gómez V., Artaza F. (Spain): A Hybridized neuro-genetic solution for controlling industrial R³ workspace, 811-824.

    This work presents a hybridized neuro-genetic control solution for R³ workspace application. The solution is based on a multi-objective genetic algorithm reference generator and an adaptive predictive neural network strategy. The trajectory calculation between two points in an R3 workspace is a complex optimization problem considering the fact that there are multiple objectives, restrictions and constraint functions which can play an important role in the problem and be in competition. We solve this problem using genetic algorithms, in a multi objective optimization strategy. Subsequently, we enhance a training algorithm in order to achieve the best adaptation of the neural network parameters in the controller which is responsible for generating the control action for a nonlinear system. As an application of the proposed hybridized control scheme, a crane tracking control is presented.

  • [3] Sánchez L., Couso I., Otero J., Palacios A. (Spain): Assessing the evolution of learning capabilities and disorders with a graphical exploratory analysis of surveys containing missing and conflicting answers, 825-838.

    The analysis of the evolution of learning with graphical maps is based on the placement of the individuals in positions that are computed on the basis of their answers to certain tests. These techniques are useful for detecting similarities between the knowledge profiles of the subjects and can also be used for assessing the acquisition of capabilities after a course. In this paper, we propose to extend some graphical exploratory analysis techniques to the case where there are missing or conflicting answers in the tests. We will also consider that either a missing or unknown answer, or a set of conflictive answers to a survey, is aptly represented by an interval or a fuzzy set. This representation causes that each individual in the map is no longer a point but a figure whose shape and size determine the coherence of the answers and whose position with respect to its neighbors determines the similarities and differences between the individuals.

  • [4] Kazienko P., Kajdanowicz T. (Poland): Base classifiers in boosting-based classification of sequential structures, 839-851.

    Boosting as a very successful classification algorithm represents a great generalization ability with appropriate ensemble diversity. It can be easily applied in the two-class classification problem. However, sequential structure prediction, in which the output is an ordered list of the labeled classes, needs to be realized by an adjusted and extended version. For that purpose the AdaBoostSeq algorithm has been introduced. It performs the multi-class classification with respect to the sequential structure of the classification target. The profile of the AdaBoostSeq algorithm is analyzed in the paper, especially its classification accuracy, using various base classifiers applied to diverse experimental datasets with comparison to other state-of-the-art methods.

  • [5] Wilk T., Woźniak M. (Poland): Combination of one-class classifiers for multiclass problems by fuzzy logic, 853-869.

    Combining classifiers, so-called Multiple Classifier Systems (MCSs), gained a lot of interest has recent years. Researchers, developed a large variety of methods in order to exploit strengths of individual classifiers. In this paper, we address the problem of how to implement a multi-class classifier by an ensemble of one-class classifiers. To improve performance of a compound classifier, different individual classifiers (which may, e.g., differ in complexity, type, training algorithm or other) can be combined and that could increase its both performance, and robustness. The model of one-class classifiers can only recognize one of the classes, therefore, it is quite difficult to produce MCSs on the basis of one-class classifiers. Thus, we introduce a new scheme for decision-making in MCSs through a fuzzy inference system. Specifically, we address two important open problems in the context: model selection and combiner training. Classifiers' outputs as supports for given classes are combined by means of a fuzzy engine. Thus, we are interested in such individual classifiers which can return support for given classes. There are no other restrictions on the used classifiers. The proposed model has been evaluated by computer experiments on several benchmark datasets in the Matlab environment. Their results prove that fuzzy combination of binary classifiers may be a valuable classifier itself. Additionally, there are indicated both some application areas of the models, and new research frontiers to be examined.

  • [6] García-Naya J. A., Dapena A., Castro P. M., Iglesia D. (Spain): DASBE: Decision-aided semi-blind equalization for MIMO systems with linear precoding, 871-882.

    Multiple-Input Multiple-Output (MIMO) digital communications standards usually acquire Channel State Information (CSI) by means of supervised algorithms, which implies loss of performance since pilot symbols do not convey information. We propose obtaining this CSI by using semi-blind techniques, which combine both supervised and unsupervised (blind) methods. The key idea consists in introducing a decision criterion to determine when the channel suffered a significant change. In such a case, transmission of pilot symbols is required. The use of this criterion also allows us to determine the time instants in which CSI has to be sent to the transmitter from the receiver through a low-cost feedback channel.

  • [7] Sedano J., Corchado E., Curiel L., Villar J. R., de la Cal E. (Spain): Detection of heat flux failures in building using a soft computing diagnostic system, 883-898.

    The detection of insulation failures in buildings could potentially conserve energy supplies and improve future designs. Improvements to thermal insulation in buildings include the development of models to assess fabric gain -heat flux through exterior walls in the building- and heating processes. Thermal insulation standards are now contractual obligations in new buildings, and the energy efficiency of buildings constructed prior to these regulations has yet to be determined. The main assumption is that it will be based on heat flux and conductivity measurement. Diagnostic systems to detect thermal insulation failures should recognize anomalous situations in a building that relate to insulation, heating and ventilation. This highly relevant issue in the construction sector today is approached through a novel intelligent procedure that can be programmed according to local building and heating system regulations and the specific features of a given climate zone. It is based on the following phases. Firstly, the dynamic thermal performance of dif\-ferent variables is specifically modeled. Secondly, an exploratory projection pursuit method called Cooperative Maximum-Likelihood Hebbian Learning extracts the relevant features. Finally, a supervised neural model and identification techniques constitute the model for the diagnosis of thermal insulation failures in building due to the heat flux through exterior walls, using relevant features of the data set. The reliability of the proposed method is validated with real datasets from several Spanish cities in winter time.

  • [8] Sánchez-Monedero J., Hervás-Martínez C., Gutiérrez P. A., Carbonero Ruz M., Ramírez Moreno M. C., Cruz-Ramírez M. (Spain): Assessing the evolution of learning capabilities and disorders with a graphical exploratory analysis of surveys containing missing nd conflicting answers, 899-912.

    Accuracy alone can be deceptive when evaluating the performance of a classifier, especially if the problem involves a high number of classes. This paper proposes an approach used for dealing with multi-class problems, which tries to avoid this issue. The approach is based on the Extreme Learning Machine (ELM) classifier, which is trained by using a Differential Evolution (DE) algorithm. Two error measures (Accuracy, $C$, and Sensitivity, S) are combined and applied as a fitness function for the algorithm. The proposed approach is able to obtain multi-class classifiers with a high classification rate level in the global dataset with an acceptable level of accuracy for each class. This methodology is evaluated over seven benchmark classification problems and one real problem, obtaining promising results.

  • [9] Fernandez-Gauna B., Lopez-Guede J. M., Zulueta E., Grana M. (Spain): Learning hose transport control with Q-learning, 913-923.

    Non-rigid physical links introduce highly nonlinear dynamics in Multicomponent Robotic Systems (MCRS), which can hardly be solved analytically. In this paper, we propose the use of reinforcement learning methods to allow the agents learn by themselves how to deal with this kind of elements, as opposed to classical control schemes. In this paper we deal with the simplest case: only one hose segment and one robot at the tip of the hose. The task is to move the hose tip to an approximate position in the space. Learning is performed and tested using a hose-MCRS simulation environment developed by our group.

  • [10] Wozniak M., Zmyslony M. (Poland): Combining classifiers using trained fuser -- analytical and experimental results, 925-934.

    Combining pattern recognition is a promising direction in designing effective classifiers. There are several approaches to collective decision-making, including quite popular voting methods where the decision is a combination of individual classifiers' outputs. The article focuses on the problem of fuser design which uses discriminants of individual classifiers to make a decision. We present taxonomy of proposed fusers and discuss some of their properties. We focus on the fuser which uses weights dependent on classifier and class number, because of a pretty low computational cost of its training. We formulate the problem of fuser learning as an optimization task and propose a solver which has its origin in neural computations. The quality of proposed learning algorithm was evaluated on the basis of several computer experiments, which were carried out on five benchmark datasets and their results confirm the quality of proposed concept.

  • [11] Savio A., Charpentier J., Termenón M., Shinn A. K., Grana M. (Spain, France, USA): Neural classifiers for schizophrenia diagnostic support on diffusion imaging data, 935-949.

    Diagnostic support for psychiatric disorders is a very interesting goal because of the lack of biological markers with sufficient sensitivity and specificity in psychiatry. The approach consists of a feature extraction process based on the results of Pearson correlation of known measures of white matter integrity obtained from diffusion weighted images: fractional anisotropy (FA) and mean diffusivity (MD), followed by a classification step performed by statistical support vector machines (SVM), different implementations of artificial neural networks (ANN) and learn vector quantization (LVQ) classifiers. The most discriminant voxels were found in frontal and temporal white matter. A total of 100% classification accuracy was achieved in almost every case, although the features extracted from the FA data yielded the best results. The study has been performed on publicly available diffusion weighted images of 20 male subjects.

  • [12] Burduk R. (Poland): The new upper bound on the probability of error in binary tree classifier with fuzzy information, 951-961.

    The paper considers the mixture of randomness and fuzziness in a binary tree classifier. This model of classification is based on fuzzy observations, the randomness of classes and the Bayes rule. In this work, we present a new upper bound on the probability of error in a binary tree classifier. The obtained error for fuzzy observations is compared with the case when observations are not fuzzy, as a difference of errors. Additionally, the obtained results are compared with the bound on the probability of error based on information energy of fuzzy events. For interior nodes of decision tree, the new bound is twice as precise as the bound based on information energy.

  • [13] Quevedo J. R., Montanées E., Ranilla J., Díaz I. (Spain): Ranked tag recommendation systems based on logistic regression, 963-977.

    This work proposes an approach to tag recommendation based on a learning system. The goal of this method is to support users of current social network systems by providing a rank of new meaningful tags for a resource. This system provides a ranked tag set and it feeds on different posts depending on the resource for which the user requests the recommendation. This research studies different approaches depending on both the posts selected to form the training set and the features with which they are represented. The performance of these approaches are tested according to several evaluation measures; one of them is proposed in this paper F1+ which takes into account the positions where the system has ranked the positive tags at the same time that it considers the cases where positive tags could not be ranked. These experiments show that this learning system outperforms certain benchmark recommenders.




6/2010

  • [1] Berkane M., Clarysse P., Magnin I. E. (Algeria, France): A neural network based summarizing method of periodic image sequences, 687-703.

    This work relates to the study of periodic events such as the ones that can be observed in biomedicine. Currently, biological processes exhibiting a periodic behaviour can be observed through the continuous recording of signals or images. Due to various reasons, cycle duration may slightly vary over time. For further analysis, it is important to be able to extract meaningful information from the mass of acquired data. This paper presents a new neural network based method for the extraction of a summarized cycle from long and massive cycle recordings. Its concept is simple and it could be naturally implemented on a hardware architecture to speed up the process. The proposed method is demonstrated on synthetic image sequences of the beating heart, and exploited as a prior in a new approach for the fast reconstruction of Magnetic Resonance Image sequences.

  • [2] Kian Ming Lim, Chu Kiong Loo, Way Soong Lim (Malaysia): Autonomous and deterministic supervised fuzzy clustering, 705-721.

    A fuzzy model based on an enhanced supervised fuzzy clustering algorithm is presented in this paper. The supervised fuzzy clustering algorithm [6] allows each rule to represent more than one output with different probabilities for each output. This algorithm implements k-means to initialize the fuzzy model. However, the main drawbacks of this approach are that the number of clusters is unknown and the initial positions of clusters are randomly generated. In this work, the initialization is done by the global k-means algorithm [1], which can autonomously determine the actual number of clusters needed and give a deterministic clustering result. In addition, the fast global k-means algorithm [1] is presented to improve the computation time. The model is tested on medical diagnosis benchmark data and Westland vibration data. The results obtained show that the model that uses the global k-means clustering algorithm [1] has higher accuracy when compared to a model that uses the k-means clustering algorithm. Besides that, the fast global k-means algorithm [1] also improved the computation time without degrading much the model performance.

  • [3] Tuček M. (Czech Republic): Transport safety, neurobehavioral disorders and medical fitness standards, 723-736.

    The effects of sleepiness, sleep loss and fatigue have been the focus of literally hundreds of studies dating back to 1896. Sleep disorders, like any other medical condition potentially affecting the safe performance of essential job functions or the safety of co-workers or the general public, require an individual assessment of the employee diagnosed with the condition to determine medical fitness for service and the necessity of any appropriate reasonable accommodations. The medical fitness assessment is a tool for maximum possible operational safety and the health and safety of all personnel in the railway industry. The article describes relevant international medical fitness standards for railway staff with special rules recommended for mental disorders, disorders of the central nervous system and use of alcohol, drugs, and other psychotropic substances.

  • [4] Popa M. C., Rothkrantz L. J. M., Datcu D., Wiggers P., Braspenning R., Shan C. (Netherlands): A comparative study of HMMs and DBNs applied to Facial Action Units Recognition, 737-760.

    From a theoretical point of view, Hidden Markov Models (HMMs) and Dynamic Bayesian Networks (DBNs) are similar, still in practice they pose different challenges and perform in a different manner. In this study we present a comparative analysis of the two spatial-temporal classification methods: HMMs and DBNs applied to the Facial Action Units (AUs) recognition problem. The Facial Action Coding System (FACS) developed by Ekman and Friesen decomposes the face into 46 AUs, each AU being related to the contraction of one or more specific facial muscles. FACS proved its applicability to facial behavior modeling, enabling the recognition of an extensive palette of facial expressions. Even though a lot has been published on this theme, it is still difficult to draw a conclusion regarding the best methodology to follow, as there is no common basis for comparison and sometimes no argument is given why a certain classification method was chosen. Therefore, our main contributions reside in discussing and comparing the relative performance of the two proposed classifiers (HMMs vs. DBNs) and also of different Region of Interest (ROI) selections proposed by us and different optical flow estimation methods. We can consider our automatic system towards AUs classification an important step in the facial expression recognition process, given that even one emotion can be expressed in different ways, fact that suggests the complexity of the analyzed problem. The experiments were performed on the Cohn-Kanade database and showed that under the same conditions regarding initialization, labeling, and sampling, both classification methods produced similar results, achieving the same recognition rate of 89% for the classification of facial AUs. Still, by enabling non-fixed sampling and using HTK, HMMs rendered a better performance of 93% suggesting that they are better suited for the special task of AUs recognition.

  • [5] Aliyeh Kazemi, M. Reza. Mehregan, Hamed Shakouri G., M. Bagher. Menhaj, Najmeh Neshat (Iran): A hierarchical artificial neural network for transport energy demand forecast: Iran case study, 761-772.

    This paper presents a neuro-based approach for annual transport energy demand forecasting by several socio-economic indicators. In order to analyze the influence of economic and social indicators on the transport energy demand, gross domestic product (GDP), population and total number of vehicles are selected. This approach is structured as a hierarchical artificial neural networks (ANNs) model based on the supervised multi-layer perceptron (MLP), trained with the back-propagation (BP) algorithm. This hierarchical ANNs model is designed properly. The input variables are transport energy demand in the last year, GDP, population and total number of vehicles. The output variable is the energy demand of the transportation sector in Million Barrels Oil Equivalent (MBOE). This paper proposes a hierarchical artificial neural network by which the inputs to the ending level are obtained as outputs of the starting levels. Actual data of Iran from 1968-2007 is used to train the hierarchical ANNs and to illustrate capability of the approach in this regard. Comparison of the model predictions with conventional regression model predictions shows its superiority. Furthermore, the transport energy demand of Iran for the period of 2008 to 2020 is estimated.

  • [6] Jiřina M., Bouchner P., Novotný S. (Czech Republic): Identification of driver's drowsiness using driving information and EEG, 773-791.

    The paper summarizes the first results of an identification of sleepy state of drivers using a complex set of outputs from simulated driving. The driving information, such as deviation from the centerline of the road and the steering wheel position as well as two-point EEG, was used. The process consists of the preprocessing of data, in fact a transformation into a form proper for classification, and a classification into one of two classes, i.e. wakefulness and drowsiness. There were two groups of drivers submitted to tests, the wakeful ones, and the drivers after serious sleep deprivation. We found that it is possible to distinguish these groups using an appropriate classifier with some rather substantial error, which can possibly be tackled by using an apt methodology.

  • [7] Fabián Z. (Czech Republic): Score correlation, 793-798.

    In this paper, we study a distribution-dependent correlation coefficient based on the concept of scalar score. This new measure of association of continuous random variables is compared by means of simulation experiments with the Pearson, Kendall and Spearman correlation coefficients.

  • [8] Contents volume 20 (2010), 799-801.

  • [9] Author's index volume 20 (2010), 803-806.




5/2010

  • [1] Yetilmezsoy K. (Turkey): Modeling studies for the determination of completely mixed, activated sludge reactor volume: Steady-state, empirical and ANN applications, 559-589.

    This paper presents an empirical model and a three-layer (7:11:1) artificial neural network (ANN) approach for the determination of completely mixed activated sludge reactor volume (CMASRV). CMASRV values were estimated by a new mathematical formulation and a three-layer ANN model for 1,000 different artificial scenarios given in a wide range of seven biological variables. The predicted results obtained from each stochastic approach were compared with the well-known steady state volume model based on mass balance equations. The computational analysis showed that the proposed empirical model and ANN outputs were obviously in agreement with the steady-state volume model and all the predictions proved to be satisfactory with a correlation coefficient of about 0.9989 and 1, respectively. The maximum volume deviations from the steady-state volume equation were recorded as only 7.17% and 6.89% for the proposed model and ANN outputs respectively. In addition to volume comparison, waste sludge mass flow rates (PX), food to mass ratios (F/M), hydraulic retention times (HRTs), volumetric organic loads (LV) and oxygen requirements (ORs) were also compared for each model, and significant points of proposed approaches were evaluated.

  • [2] Snášel V., Platoš J. , Krömer P., Abraham A., Ouddane N., Húsek D. (Czech Republic): Interleaver optimization using population based metaheuristics, 591-608.

    Since their appearance in 1993, first approaching the Shannon limit, turbo codes have given a new direction in the channel encoding field, especially since they have been adopted for multiple norms of telecommunications such as deeper communication. A robust interleaver can significantly contribute to the overall performance a turbo code system. Search for a good interleaver is a complex combinatorial optimization problem. In this paper, we present genetic algorithms and differential evolution, two bio-inspired approaches that have proven the ability to solve non-trivial combinatorial optimization tasks, as promising optimization methods to find a well-performing interleaver for large frame sizes.

  • [3] Votruba Z., Novák M. (Czech Republic): Alliance approach to the modeling of interfaces in complex heterogeneous objects, 609-619.

    Fundamentals of the theory of system alliances are briefly reviewed. An accent is put on interfaces (IFs). The model of IFs consisting of a pair of finite deterministic automata sharing a part of their internal state space is introduced. The presented model of alliance interface can be successfully implemented for the study of typical phenomena in complex heterogeneous objects with a significant degree of uncertainty.

  • [4] Boostani R., Dehzangi O., Jarchi D., Zolghadri Mansoor J. (Iran): An efficient pattern classification approach: Combination of weighted LDA with weighted nearest neighbor, 621-635.

    Linear discriminant analysis (LDA) is a versatile method in all pattern recognition fields but it suffers from some limitations. In a multi-class problem, when samples of a class are far from other classes samples, it leads to bias of the whole decision boundaries of LDA in favor of the farthest class. To overcome this drawback, this study is aimed at minimizing this bias by redefining the between- and within-class scatter matrices via incorporating weight vectors derived from Fisher value of classes pairs. After projecting the input patterns into a lower-dimensional space in which the class samples are more separable, a new version of nearest neighbor (NN) method with an adaptive distance measure is employed to classify the transformed samples. To speed up the adaptive distance routine, an iterative learning algorithm that minimizes the error rate is presented. This efficient method is applied to six standard datasets driven from the UCI repository dataset and test results are evaluated from three aspects in terms of accuracy, robustness, and complexity. Results show the supremacy of the proposed two-layer classifier in comparison with the combination of different versions of LDA and NN methods from the three points of view. Moreover, the proposed classifier is assessed in the noisy environment of those datasets and the achieved results confirm the high robustness of the introduced scheme when compared to others.

  • [5] Zare A., Zolghadri Jahromi M., Boostani R. (Iran): An adaptive-distance artificial immune recognition system, 637-650.

    The artificial Immune Recognition System (AIRS) algorithm inspired by a natural immune system makes use of the training data to generate memory cells (or prototypes). These memory cells are used in the test phase to classify unseen data using the K-nearest neighbor (K-NN) algorithm. The performance of the AIRS algorithm, similar to other distance-based classifiers, is highly dependent on the distance function used to classify a test instance. In this paper, we present a new version of the AIRS algorithm named Adaptive Distance AIRS (AD-AIRS) that uses an adaptive distance metric to improve the generalization accuracy of the basic AIRS algorithm. The adaptive distance metric is based on assigning weights to the evolved memory cells. The weights of memory cells are used in the test phase to classify test instances. Apart from this, the AD-AIRS algorithm uses the concept of clustering to modify the way that memory cells are generated. Each memory cell represents a group of similar instances (or antigens). A subset of the UCI datasets is used to evaluate the effectiveness of the proposed AD-AIRS algorithm in comparison with the basic AIRS. Experimental results show that the AD-AIRS achieves higher accuracy with a fewer number of memory cells when compared with the basic AIRS algorithm.

  • [6] Lin L., Gang D. (China): A multiple classification method based on the cloud model, 651-666.

    Based on the randomness and fuzziness of the cloud model during the transformation from the qualitative concept to the quantitative numerical value, with the theory that any data distribution can be decompounded into several normal distributions, this paper puts forward a method of multi-classification based on the cloud model. By this method, multiple classification is transformed to a superposed cloud model with training samples as the cloud expectation, while the test samples are regarded as the `cloud droplets', and their classifications of membership degree in a cloud model can be calculated. Considering the effect of the number of training samples on the membership degree, the cloud model is weighted by the ratio of the total number of training samples to the number of training samples in a single class so that the data distribution of the samples can be balanced. The formula of multiple classification based on the cloud model has the structure identical to that of Support Vector Machines, and the hyper entropy in cloud models exerts similar punishment on the noise samples just like the loose coefficients in Support Vector Machines; therefore, the reasonability of the method is theoretically proved. Compared with Support Vector Machine, the method discussed in this paper does not require any large-scale quadratic programming, thus the algorithm of the method is simpler. Last but not the least, five types of data distribution samples are selected for the comparative experiment, and comparison is made with four other classification methods; the result shows that the accuracy and stability of the algorithm is high, and its implementations on the high dimensional multiple classifications are especially satisfactory.

  • [7] Ahmad I., Shah A., Khan A. N. (Pakistan): Application of neural network model for the prediction of shear strength of reinforced concrete beams, 667-686.

    This paper evaluates the feasibility of using an Artificial Neural Network (ANN) model for estimating the nominal shear capacity of Reinforced Concrete (RC) beams against diagonal shear failure subjected to shear and flexure. A feedforward back-propagation ANN model was developed utilizing 622 experimental data points of RC beams, which include 111 deep beams data and 20 beams tested for low longitudinal steel ratios. The ANN model was trained on 70% of the data and then it was validated using the remaining 30% data (new data were not used for training). The trained ANN model was compared with three existing approaches, including the American Concrete Institute (ACI) code. The ANN model predictions when compared to the experimental data were very favorable, regarding also the other approaches. The prediction of ANN model was also checked for size effect and deep beams separately. The ANN model was found to be very robust in all situations. The safe form of ANN model was also derived and compared with the design equations of the three methods.

4/2010

  • [1] Guang-ming Xian, Bi-qing Zeng (China): An effective and novel fault diagnosis technique based on EMD and SVM, 427-439.

    An effective and novel roller bearing fault diagnosis technique based on empirical mode decomposition (EMD) energy entropy and support vector machine (SVM) is put forward in this article. The vibration signal of roller bearing is decomposed by EMD and the first 5 intrinsic mode function (IMF) components are obtained. SVM served as a fault diagnosis classifier and the extracted energy features of the first 5 IMFs are taken as network input vectors, and then the fault bearing and the normal bearing can be distinguished. An technique for fault of roller bearing by SVM is evaluated against a series of fault diagnosis methods that are widely used in machinery, with particular regard to the effect of training set size on fault diagnosis accuracy. We trained the SVM using RBF kernel function. We compare our experimental results with the existing results given by SMO and SVM-light algorithms. It can be seen that the fault diagnosis method based on SVM-light is superior to that based on SMO in diagnosis accuracy of roller bearing. In addition to the SVM, the same datasets were classified using RBF NN and Hopfield NN. The experimental results show that the technique of support vector machine based on EMD energy entropy has higher fault diagnosis ability.

  • [2] Harrag F., Hamdi-Cherif A., El-Qawasmeh E. (Algeria, Saudi Arabi, Jordan): Performance of MLP and RBF neural networks on arabic text categorization using SVD, 441-459.

    Text categorization is based on the idea of content-based texts clustering. An Artificial Neural Network (ANN) or simply Neural Network (NN) classifier for Arabic texts categorization is proposed. The Singular Value Decomposition (SVD) is used as preprocessor with the aim of further reducing data in terms of both size and dimensionality. Indeed, the use of SVD makes data more amenable to classification and the convergence training process faster. Specifically, the effectiveness of the Multilayer Perceptron (MLP) and the Radial Basis Function (RBF) classifiers are implemented. Experiments are conducted using an in-house corpus of Arabic texts. Precision, recall and F-measure are used to quantify categorization effectiveness. The results show that the proposed SVD-Supported MLP/RBF ANN classifier is able to achieve high effectiveness. Experimental results also show that the MLP classifier outperforms the RBF classifier and that the SVD-supported NN classifier is better than the basic NN, as far as Arabic text categorization is concerned.

  • [3] Űneş F. (Hatay-Turkey): Dam reservoir level modeling by neural network approach: A case study, 461-474.

    Prediction of reservoir level fluctuation is important in the operation, design, and security of dams. In this paper, Artificial Neural Networks (ANN) is used for modeling. In such modeling approaches, it is possible to determine dam reservoir level and water balance (budget) by taking the monthly average precipitation and needed parameters into consideration. The basic data are available for over 29 years at the Tahtakőprű Dam in the southeast Mediterranean region of Turkey. As a sub-approach of ANN, a multi layer perceptron (MLP) is used. Bayesian regularization back-propagation training algorithm is employed for optimization of the network. MLP results are compared with the results of conventional multiple linear regression (MLR) and autoregressive (AR) models. The comparison shows that the ANN model provides better performance than the mentioned models in reservoir level estimation.

  • [4] Bravo C., L'Huillier G., Luis Lobato J., Weber R. (Chile): Probability estimation for multiclass problems combining SVMs and neural networks, 475-489.

    This paper addresses the problem of probability estimation in Multiclass classification tasks combining two well-known data mining techniques: Support Vector Machines and Neural Networks. We present an algorithm which uses both techniques in a two-step procedure. The first step employs Support Vector Machines within a One-vs-All reduction from multiclass to binary approach to obtain the distances between each observation and the Support Vectors representing the classes. The second step uses these distances as inputs for a Neural Network, built with an entropy cost function and softmax transfer function for the output layer where class membership is used for training. Consequently, this network estimates probabilities of class membership for new observations. A benchmark using different databases demonstrates that the proposed algorithm is highly competitive with the most recent techniques for multiclass probability estimation.

  • [5] Fabian Z. (Czech Republic): Spectral estimation of non-Gaussian time series, 491-499.

    Based on the concept of the scalar score of a probability distribution, we introduce a concept of a scalar score of time series and propose to characterize a non-Gaussian time series by spectral density of its scalar score.

  • [6] Yue B., Liu H., Abraham A., Badr Y. (USA, France): A multi-swarm synergetic optimizer for multi-knowledge extraction using rough set, 501-517.

    Finding reducts is one of the key problems in the increasing applications of rough set theory, which is also one of the bottlenecks of the rough set methodology. The population-based reduction approaches are attractive to find multiple reducts in the decision systems, which could be applied to generate multi-knowledge and to improve decision accuracy. In this paper, we design a multi-swarm synergetic optimization algorithm (MSSO) for rough set reduction and multi-knowledge extraction. It is a multi-swarm based search approach, in which different individual trends to be encoded to different reduct. The approach discovers the best feature combinations in an efficient way to observe the change of positive region as the particles proceed throughout the search space. The performance of our approach is evaluated and compared with Standard Particle Swarm Optimization (SPSO) and Genetic Algorithms (GA). Empirical results illustrate that the approach can be applied for multiple reduct problems and multi-knowledge extraction effectively.

  • [7] Lee J.-S., Park C. H.(Republic of Korea, Switzerland): Global optimization of radial basis function networks by hybrid simulated annealing, 519-537.

    This paper presents a global optimization method of radial basis function networks. In the proposed method, stochastic search by simulated annealing is combined with a local search technique in order to perform global optimization of the network parameters with enhanced convergence speed. Its convergence property is proved mathematically. Experimental results demonstrate that the proposed method improves the performance of the networks over the conventional local and global training methods and reduces influence of the initial parameter values on the final results.

  • [8] Faber J., Pěkný J., Pieknik R., Tichý T., Faber V., Bouchner P., Novák M. (Czech Republic): Simultaneous recording of electric and metabolic brain activity, 539-557.

    Electric and metabolic brain activities were controlled for 49 persons by means of electroencephalography (EEG) and near infrared spectroscopy (NIRS) during different physiological and psychological states, e.g. eyes open or closed, hyperpnoea (HV) or apnoea (AP) and during calculation (addition of one- (CAL.S.) or two-digit numbers (CAL C.)) or Raven's test - set A (RAV A) or more difficult set C (RAV C) or TAT. Both diagnostic methods confirm one another. But sometimes there are more expressive changes in one method than in another, e.g. during HV it is a more prominent deoxyhemoglobin (CO2Hb) increase than an alpha frequency band increase in EEG curves. On the other side, very marked alpha frequency decreasing in EEG during eyes open is followed by only a weak oxyhemoglobin (O2Hb) rise. Synchronous activities are clearly visible during psychic test: increasing O2Hb in NIRS and increasing delta frequency band in EEG spectrum. Decreasing alpha with increasing theta activity during relaxation and somnolence is accompanied by decreasing O2Hb and prominent increasing CO2Hb. Inter individually differences are often but not big. We suggest both methods are good tools for attention and psychic states control.

3/2010

  • [1] Editorial, 261-264.
  • [2] Zhiyong Zhang, Shiguo Lian, Qingqi Pei, Jiexin Pu (China): Fuzzy risk assessments on security policies for digital rights management, 265-284.

    In multimedia consuming, Digital Rights Management (DRM) is the important means to confirm the benefits of both digital contents/services providers and consumers. To keep the DRM system running in order, risk management should be adopted, which identifies and assesses the DRM system's security level. Now, the legitimate sharing of copyrighted digital content is still an open issue, which faces severe risks of propertied assets circumvention and copyright infringements. In this paper, we try to highlight a multi-disciplinary method for all-around examinations on risks to digital assets in the contents sharing scenario. The method is a qualitative and quantitative fuzzy risk assessment, which is used for estimating a novel concept called Risk-Controlled Utility (RCU) in DRM. Then, we emphasize an application case of the emerging trusted computing policy, and analyze the influences of different content sharing modes. Finally, we address a business model with some simulation results. Comparison with other methods shows that the fusion of qualitative and quantitative styles cannot only evaluate the RCU with uncertain risk events effectively, but also provide accurate assessment data for the security policies of DRM.

  • [3] Xianglian Xue, Qiang Zhang, Xiaopeng Wei, Ling Guo, Qian Wang (China): A digital image encryption algorithm based on DNA sequence and multi-chaotic maps, 285-296.

    This paper presented a new image encryption algorithm. The algorithm includes two steps: first, by using Cubic map and wavelet function to produce the 2D chaotic sequences to scramble the location of pixel points from the image, then using DNA sequence and chaotic sequence produced by Logistic chaotic map to disturb the gray of the pixel points from image. The experimental results and security analysis show that our algorithm can get good encryption effect, has widest secret key's space, strong sensitivity to secret key, and has the ability of resisting exhaustive attack and statistic attack.

  • [4] Sajedi H., Jamzad M. (Iran): Evolutionary rule generation for signature-based cover selection steganography, 297-316.

    A novel approach for selecting proper cover images in steganography is presented in this paper. The proposed approach consists of two stages. The first stage is an evolutionary algorithm that extracts the signature of cover images against stego images in the form of fuzzy if-then rules. This algorithm is based on an iterative rule learning approach to construct an accurate fuzzy rule base. The rule base is generated in an incremental way by optimizing one fuzzy rule at a time using an evolutionary algorithm. In the second stage of the proposed approach, the fuzzy rules generated in the first stage are used for selecting suitable cover images for steganography. We applied our approach to some state-of-the-art steganography techniques and validated it using an image database. The results indicate that a secret message can be securely embedded in selected cover images. Therefore, we can apply the proposed evolutionary fuzzy algorithm, as an intelligent rule generation approach, to select the appropriate cover images from an image database and use them to have more secure steganography.

  • [5] Wanyu Deng, Lin Chen (China): Color image watermarking using regularized extreme learning machine, 317-330.

    In this paper, a real-time watermarking scheme based on Regularized Extreme Learning Machine (RELM) is proposed. Using the information provided by the reference positions, RELM can be trained at the embedding procedure and watermark is adaptively embedded into the blue channel of the original image by considering the human visual system. Due to the extreme training speed (always hundreds of times faster than BP neural network and Support Vector Machine (SVM)) and good generalized performance, the trained RELM can exactly extract the watermark from the watermarked image against image processing attacks within very short time, and this makes this method applicable in real-time environment. Extensive experimental results illustrate that our technique outperforms Kutter's and Yu's method against simple and multiple attacks.

  • [6] Ebrahim G. A., Younis A. A. (Egypt, USA): A two-stage VoIP spam identification framework, 331-358.

    Identifying a VoIP call as SPAM based on call characteristics is an important issue that has never been studied before. Most of the studies of VoIP SPAM impose the whole burden on the callee to judge SPAM calls. In other words, the accuracy of the identification process is totally based on the callee identifying the call as SPAM, which is questionable and not reliable. In this paper, a two-stage VoIP SPAM identification framework is introduced. The first stage is a pre-call identification process, which uses a set of parameters about the call that can be collected before allowing the call to go through. The second stage is a post-call identification process that uses other parameters that can be collected during/after the call. The first stage provides a pre-call evaluation score of the call, while the second stage further tunes this score. In the proposed framework, the decision of identifying VoIP SPAM calls is based on several uncertain parameters that represent meta-data of VoIP calls. These parameters include call duration, amount of exchanged information in each direction, and calling pattern. In this study, the potential set of parameters that can be used to identify VoIP SPAM are investigated. A set of rules is used in addition to any prior evaluation of the caller to provide the pre-call score. Then, a fuzzy-logic controller is developed to identify VoIP SPAM in the second stage. An augmented ongoing tuning strategy is adopted where callee feedback, if any, is taken into account to further tune the identification process. Simulation studies are carried out to demonstrate the effectiveness of the two-stage approach in identifying VoIP SPAM based on the proposed framework.

  • [7] Young Jae Lee, Ajith Abraham, Dong Hwa Kim (Korea, USA): 3D object recognition using octree model and fast search algorithm, 359-369.

    This paper presents a new approach to 3D object recognition by using an Octree model library (OML) I, II and fast search algorithm. The fast search algorithm is used for finding the 4 pairs of feature points to estimate the viewing direction uses on effective two level database. The method is based on matching the object contour to the reference occluded shapes of 49, 118 viewing directions. The initially bestmatched viewing direction is calibrated by searching for the 4 pairs of feature points between the input image and the image projected along the estimated viewing direction. At this point, the input shape is recognized by matching it to the projected shape. The computational complexity of the proposed method is shown to be O(n^2) in the worst case, and that of the simple combinatorial method of O(m^4,n^2), where n and m denote the number of feature points of the 3D model object and the 2D object, respectively.

  • [8] Muda A. K., Shamsuddin S. M., Abraham A. (Malaysia, Norway): Improvement of authorship invarianceness for individuality representation in writer identification, 371-387.

    Writer Identification (WI) is one of the areas in pattern recognition that have created a center of attention for many researchers to work in. Recently, its main focus is in forensics and biometric application, e.g. writing style can be used as biometric features for authenticating individuality uniqueness. Existing works in WI concentrate on feature extraction and classification task in order to identify the handwritten authorship. However, additional steps need to be performed in order to have a better representation of input prior to the classification task. Features extracted from the feature extraction task for a writer are in various representations, which degrades the classification performance. This paper will discuss this additional process that can transform the various representations into a better representation of individual features for Individuality of Handwriting, in order to improve the performance of identification in WI.

  • [9] Ahmadian K., Gavrilova M. (Canada): Transiently chaotic associative network for fingerprint image analysis, 389-403.

    This paper presents a new technique for fingerprint image matching in biometric security applications, based on the hybrid of Neural Network and Delaunay Triangulation methodology. The Delaunay triangulation of the minutiae set is transformed to a set of points in the discretized space using duality. This translation results in a sampling method be acquiring which the system tolerates displacement and noise of the input image. Finally, Transiently Chaotic Associative Network (TCAN) is used to learn the obtained pattern. Experimental results show a significant improvement in the False Rejection Rate over both the traditional Delaunay Triangulation based approach and direct Neural Network application.

  • [10] Kwang-Baek Kim, Hae-Jung Lee, Doo Heon Song, Young Woon Woo (Korea): Extracting fascia and analysis of muscles from ultrasound images with FCM-based quantization technology, 405-416.

    In this paper, we propose a novel method to measure the muscle thickness from ultrasound images of lumbar region in order to diagnose low back pain effectively. Images used in this study were obtained by muscle endurance tests for analyzing low back pain and muscle contraction patterns. We measure the thickness of the third muscle layer of lumbar spine, which is the most developed one in that region, and the measuring point is the center of ultrasound image. We apply Fuzzy C-Means (FCM) based quantization in extracting fascia from ultrasound images. FCM quantizer first analyzes the distribution of intensity from images and then makes clusters of similar intensity based on the distance from the center point. Our FCM based quantizer in conjunction with end-in search stretching algorithm overcomes the disadvantage of popular ART2 based quantizer that is weak in accuracy of extracting fascia when the distribution of the intensity is diverse. Based on our experiment, the proposed method is sufficiently competitive with that of human experts in measuring thickness of lumbar region muscles, thus it could be used as an auxiliary computer aided system for medical specialists.

  • [11] S. R. Kannan, S. Ramathilagam, R. Pandiyarajan, Shiguo Lian, A. Sathya (Taiwan, India, China): Improved fuzzy clustering segmentation for medical images, 417-426.

    The purpose of this paper is to develop some effective robust fuzzy c-means methods for segmentation of Brain Medical Images and Dynamic Contrast-Enhanced Breast Magnetic Images (DCE-BMRI). Segmentation is a difficult task and challenging problem in the brain and breast medical images for diagnosing Breast and Brain cancer related diseases before the image goes for treatment plan. This paper presents three new effective fuzzy clustering techniques: Robust KFCM (Kernel Fuzzy C-Means) with spatial information, Effective Robust FCM based Kernel function, Modified fuzzy c-means algorithm with weight Bias Estimation. In experiments, the presented methods are compared with other reported methods. Experimental results on both breast and brain MR images show that the proposed algorithms have better performance than the standard algorithms. Thus, the proposed method is capable of dealing with the intensity in-homogeneities and noised image effectively.




2/2010

  • [1] Tanikić D., Manić M., Devedžić G., Ćojbašić Ž. (Serbia): Modelling of the temperature in the chip-forming zone using artificial intelligence techniques, 171-187.

    Heat generation in the cutting zone occurs as a result of the work done in metal cutting. In this study, in order to measure the temperature generated in the chip-forming zone, numerous experiments were carried out for different cutting regimes. During these experiments, the chip's top temperature was measured using an infrared camera. Collected data were analyzed, and temperature dependence on various cutting regimes was formulated. After that, measured data were modelled using the various techniques: response surface methodology, various types of artificial neural networks and neuro-fuzzy system. The accuracy of the proposed models is presented as well as their suitability for the considered problem. Finally, the system for the adaptive control of the cutting temperature, based on the proposed models, is presented.

  • [2] Panchi Li, Kaoping Song, Erlong Yang (China): Model and algorithm of neural networks with quantum gated nodes, 189-206.

    On the basis of analyzing the principles of the quantum rotation gates and quantum controlled-NOT gates, an improved design for CNOT gated quantum neural networks model is proposed and a smart algorithm for it is derived in our paper, based on the gradient descent algorithm. In the improved model, the input information is expressed by the qubits, which, as the control qubits after being rotated by the rotation gate, control the qubits in the hidden layer to reverse. The qubits in the hidden layer, as the control qubits after being rotated by the rotation gate, control the qubits in the output layer to reverse. The networks output is described by the probability amplitude of state |1> in the output layer. It has been shown in two application examples of pattern recognition and function approximation that the proposed model is superior to the standard error back-propagation networks with regard to their convergence rate, number of iterations, approximation ability, and robustness.

  • [3] Moallem P., Ayoughi S. A. (Iran): Improving backpropagation via an efficient combination of a saturation suppression method and momentum term, 207-222.

    The gradient descent backpropagation (BP) algorithm that is widely used for training MLP neural networks can retard convergence due to certain features of the error surface like the local minimum and the flat spot. Common promoting methods, such as applying momentum term and using dynamic adaptation of learning rates, can enhance the performance of BP. However, saturation state of hidden layer neurons, which is the cause of some flat spots on the error surface, persists through such accelerating methods. In this paper, we propose a grading technique to gradually level off the potential flat spots into a sloping surface in a look-ahead mode; and thereby progressively renew saturated hidden neurons. We introduce symptoms indicating saturation state of hidden nodes. In order to suppress the saturation, we added a modifying term to the error function only when saturation is detected. In normal conditions, the improvement made to the learning process is adding a momentum term to the weight correction formula. We have recorded remarkable improvements in a selection of experiments.

  • [4] Temel T. (Turkey): A new digital cochlea model neuro-spike representation of auditory signals and its application to classification of bat-like biosonar echoes, 223-239.

    For an improved neuro-spike representation of auditory signals within cochlea models, a new digital ARMA-type low-pass filter structure is proposed. It is compared to more conventional AR-type counterpart on a classification of biosonar echoes, in which echoes from various tree species insonified with a bat-like chirp call are converted to biologically plausible feature vectors. Next, parametric and non-parametric models of the class-conditional densities are built from the echo feature vectors. The models are deployed in single-shot and sequential-decision classification algorithms. The results indicate that the proposed ARMA filter structure offers an improved single-echo classification performance, which leads to faster sequential-decision making than its AR-type counterpart.

  • [5] Svítek M., Moos P., Votruba Z. (Czech Republic): Towards information circuits, 241-247.

    A trial of analogies utilization among electrical, mechanical and information circuits is presented. The concepts of Information Power and significant proximity of the measure of information and knowledge could enable upgrading these analogies for solving important tasks from the area of Systems Engineering. This attempt seems to be attractive, as it could help in using the well-established and proved methodologies from the classical areas of electricity or mechanics.

  • [6] Faber J. (Czech Republic): Biofeedback and brain activity, 249-260.

    Biofeedback is a treatment technique in which people are trained to improve their physiological functions by using different signals from their own bodies, e.g. from skin, heart (ECG), muscles (EMG), brain (EEG) etc. Psychotherapeutists use it to decrease intrapsychic tension in anxious and depressive patients and epileptics or learn to relax boys who suffered from attention deficit and hyperactivity disorders. The main system for consciousness (thalamocortical reverberation circuit) generates whole brain electromagnetic frequencies permanently (1-30 Hz = EEG activity). But we choose a specific frequency band, e.g. SMR (Sensory Motor Rhythm = 13-18 Hz) and these SMR episodes are rewarded by success in a simultaneously watched TV game. SMR is then repeated still more often and brings into electrogenesis and into psyche tendency its own property, which is motor inhibition and increasing attention. This is the aim of the therapeutical learning process.

1/2010

  • [1] Editorial, 1-2.
  • [2] Guest Editorial, 2-5.
  • [3] Kreinovich V., Perfilieva I. (USA, Czech Republic): A broad prospective on fuzzy transforms: From gauging accuracy of quantity estimates to gauging accuracy and resolution of measuring physical fields, 7-25.

    Fuzzy transform is a new type of function transforms that has been successfully used in different applications. In this paper, we provide a broad prospective on fuzzy transform. Specifically, we show that fuzzy transform naturally appears when, in addition to measurement uncertainty, we also encounter another type of localization uncertainty: that the measured value may come not only from the desired location x, but also from the nearby locations.

  • [4] Kupka J., Tomanová I. (Czech Republic): Some extensions of mining of linguistic associations, 27-44.

    This paper is a contribution to the theoretical foundations of data mining. More precisely, we contribute to a part of data mining allowing us to search for associations among attributes that can be expressed in the form of natural language sentences. The theoretical background and also a method for mining such associations was published recently in [V. Novák et al., Mining pure linguistic associations from numerical data, Int. Journal of Approximate Reasoning 48 (2008), 4 -- 22]. We elaborated other mathematical representations of the model presented in the mentioned paper in order to extend its applicability.

  • [5] Bohacik J. (Slovak Republic): Discovering fuzzy rules in databases with linguistic variable elimination, 45-61.

    A group of fuzzy IF-THEN rules is belonging to one of the most popular, most effective, and user-friendliest knowledge representations. For this reason, extraction of these rules is becoming a more-and-more important part of the Data Mining stage in the Knowledge Discovery in Databases Process. In this paper, a direct algorithm for extracting fuzzy IF-THEN rules on the basis of linguistic variable elimination is described. The algorithm is implemented within a designed object-oriented software library Fuzzy Rule Miner. Besides the introduced algorithm, it implements two algorithms for fuzzy rule extraction based on using fuzzy decision trees of ID3 kind. An essential precondition for comparing the implemented algorithms and for verifying the legitimacy of the introduced algorithm is performance of experiments. The goal of experiments is to take in the behavior of algorithms on testing databases from the UCI Repository of Machine Learning Databases and to make comparisons of algorithms with one another. According to the conducted experiments, the introduced algorithm achieves high accuracy levels of discovered knowledge. The paper also contains a classification of rules and a specification of the Fuzzy Rule Discovery in Databases Process.

  • [6] Petrík M., Sarkoci P. (Czech Republic, Slovak Republic): Zero-reconstructible triangular norms as universal approximators, 63-67.

    This paper is inspired by recent results [15, 16] which have shown that a multiplicative generator of a strict triangular norm can be reconstructed from the first partial derivatives of the triangular norm on the segment {0} x [0,1]. The strict triangular norms to which this method is applicable have been called zero-reconstructible triangular norms. This paper shows that every continuous triangular norm can be approximated (with an arbitrary precision) by a zero-reconstructible one, and thus substantiates the significance of this subclass of strict triangular norms.

  • [7] Jágr V., Komorníková M., Mesiar R. (Slovak Republic): Conditioning stable copulas, 69-79.

    Copulas stable under univariate conditioning are studied. Limit approach to construction of conditioning stable copulas is introduced and illustrated. In the class of Archimedean copulas, Clayton copulas are shown to be the only conditioning stable copulas. Conditioning stable singular copulas are also discussed and examples of non-Archimedean absolutely continuous copulas which are conditioning stable are given.

  • [8] Bacigál T., Juráňová M., Mesiar R. (Slovak Republic): On some new constructions of Archimedean copulas and applications to fitting problems, 81-90.

    Several constructions of additive generators of binary Archimedean copulas are introduced and discussed. Extension to general Archimedean copulas is also included. Applications to the fitting of copulas to real data are given and examplified.

  • [9] de Andrés R., García-Lapresta J. L. (Spain): An endogenous human resources selection model based on linguistic assessments, 91-111.

    This paper proposes an endogenous human resources selection process by using linguistic information from a competency management perspective. We consider different sets of appraisers taking part in the evaluation process, having a different knowledge about the candidates that are being evaluated. Then, appraisers can express their assessments in different linguistic domains according to their knowledge. The proposed method converts each linguistic label into a fuzzy set on a common domain. Candidates are ranked by using different aggregation operators in order to allow the management team to make a final decision.

  • [10] Fuerst K. (Austria): Applying fuzzy models in rating systems, 113-124.

    In the following paper, the use of fuzzy models in qualitative rating systems is analyzed in detail. The author works in an Austrian finance institution. There are at the moment two rating systems in use. The main purpose of such a rating system is to analyze company ratios to calculate a rating score, which is a measure for the financial situation and rigidity of a company. The first one is a solely hard fact rating system based on the Quicktest by Kralicek. The second one uses self-organizing maps and neural networks to calculate a rating classification and also offers the possibility to dispose personal appraisal in the calculation process.

    The following work examines the application spectrum of fuzzy logic and fuzzy models in soft-fact rating systems.

    We show that the use of fuzzy models in rating systems enables visualization of additional knowledge and offers the possibility to enhance the influence of a company's soft fact rating to the overall rating.

  • [11] Bebčáková I., Talašová J., Pavlačka O. (Czech Republic): Fuzzification of Choquet integral and its application in multiple criteria decision making, 125-137.

    A common approach in the multiple criteria decision making is to obtain the overall evaluation by aggregating the partial evaluations. For this, a member of a large family of aggregation operators is used. Many of these operators commonly employed in decision making (weighted average, ordered weighted average, minimum, maximum, ...) can be used only when criteria are independent. On the other hand, the Choquet integral, a generalization of the aforementioned operators, can be used even when some interactions between criteria occur. We present a fuzzified Choquet integral capable of dealing not only with fuzzy partial evaluations (first level fuzzification), but also with fuzzy weights (second level fuzzification). We also provide an effective way to evaluate the fully fuzzified integral, which allows its straightforward application to decision making problems with inherent uncertainty.

  • [12] Galar M., Bustince H., Fernandez J., Sanz J., Beliakov G. (Spain, Australia): Fuzzy entropy from weak fuzzy subsethood measures, 139-158.

    In this paper, we propose a new construction method for fuzzy and weak fuzzy subsethood measures based on the aggregation of implication operators. We study the desired properties of the implication operators in order to construct these measures. We also show the relationship between fuzzy entropy and weak fuzzy subsethood measures constructed by our method.

  • [13] Poláková R. (Czech Republic): A variant of competitive differential evolution algorithm with exponential crossover, 159-169.

    The differential evolution (DE) algorithm is a powerful population-based stochastic technique to search for global optimum in the continuous search space. Success of DE algorithm strongly depends on choosing its parameters. The competition in differential evolution was shown to be an efficient instrument to avoid time-consuming process of tuning control parameters. A new variant of competitive DE algorithm, called BEBERAN, was proposed and tested on benchmark functions at four levels of the search space dimension. The BEBERAN was compared with the most promising competitive variant, DEBR18. BEBERAN, in contrast to DEBR18, includes in addition the exponential crossover.