Contents of Volume 27 (2017)

1/2017 2/2017 3/2017 4/2017 5/2017

6/2017

  • [1] Fabera V., Musil T., Rada J. (CZ)
    The first hardware MSC algorithm implementation, 541-555

      Full text     DOI: http://dx.doi.org/10.14311/NNW.2017.27.029

    Abstract: This paper describes the first attempt of hardware implementation of Multistream Compression (MSC) algorithm. The algorithm is transformed to series of Finite State Machines with Datapath using Register-Transfer methodology. Those state machines are then implemented in VHDL to selected FPGA platform. The algorithm utilizes a special tree data structure, called MSC tree. For storage purpose of the MSC tree a Left Tree Representation is introduced. Due to parallelism, the algorithm uses multiple port access to SDRAM memory.

  • [2] Xue X., Yang X., Li P., (China)
    Application of a probabilistic neural network for liquefaction assessment, 557-567

      Full text     DOI: http://dx.doi.org/10.14311/NNW.2017.27.030

    Abstract: This paper presents a hybrid probabilistic neural network (PNN) and particle swarm optimization (PSO) techniques to predict the soil liquefaction. The PSO algorithm is employed in selecting the optimal smoothing parameter of the PNN to improve the forecasting accuracy. Seven parameters such as earthquake magnitude, normalized peak horizontal acceleration at ground surface, standard penetration number, penetration resistance, relative compaction, mean grain diameter and groundwater table are selected as the evaluating indices. The predictions from the PSO-PNN model were compared with those from two models: back-propagation neural network (BPNN) model and support vector machine (SVM) model. The study concluded that the proposed PSO-PNN model can be used as a reliable approach for predicting soil liquefaction.

  • [3] Temel T. (Turkey)
    A New Classification Algorithm: Optimally Generalized Learning Vector Quantization (OGLVQ), 569-576

      Full text     DOI: http://dx.doi.org/10.14311/NNW.2017.27.031

    Abstract: We present a new Generalized Learning Vector Quantization classifier called Optimally Generalized Learning Vector Quantization based on a novel weightupdate rule for learning labeled samples. The algorithm attains stable prototype/weight vector dynamics in terms of estimated current and previous weights and their updates. Resulting weight update term is then related to the proximity measure used by Generalized Learning Vector Quantization classifiers. New algorithm and some major counterparts are tested and compared for synthetic and publicly available datasets. For both the datasets studied, it is seen that the new classifier outperforms its counterparts in training and testing with accuracy above 80% its counterparts and in robustness against model parameter varition.

  • [4] Mohamed Kh.Sh., W. Wu, Y. Liu (China)
    A modified higher-order feed forward neural network with smoothing regularization, 577-59

      Full text     DOI: http://dx.doi.org/10.14311/NNW.2017.27.032

    Abstract: This paper proposes an offline gradient method with smoothing L1/2 regularization for learning and pruning of the pi-sigma neural networks (PSNNs). The original L1/2 regularization term is not smooth at the origin, since it involves the absolute value function. This causes oscillation in the computation and difficulty in the convergence analysis. In this paper, we propose to use a smooth function to replace and approximate the absolute value function, ending up with a smoothing L1/2 regularization method for PSNN. Numerical simulations show that the smoothing L1/2 regularization method eliminates the oscillation in computation and achieves better learning accuracy. We are also able to prove a convergence theorem for the proposed learning method.

  • [5] Brandejsky T.(CZ)
    Influence of (p)RNGs onto GPA-ES behaviors , 593-605

      Full text     DOI: http://dx.doi.org/10.14311/NNW.2017.27.033

    Abstract: The main aim of this paper is to investigate if the evolutionary algorithms (EAs) can be influenced by Random Number Generators (RNGs) and pseudo Random Number Generators (pRNGs) and if different evolutionary operators applied within EAs requires different features of RNGs and pRNGs. Speaking both about RNGs and pRNGs, the abbreviation (p)RNGs will be used. This question is significant especially if genetic programming is applied to symbolic regression task with the aim to produce human expert comparable results because such task requires massive computations. Experiments were performed on GPA-ES algorithm combining genetic programming algorithm (GPA) for structure development and evolutionary strategy (ES) algorithm for parameter optimization. This algorithm is described bellow and it applies extended scale of different evolutionary operators (additional individuals generating, symmetric crossover, mutations, and one point crossover). These experiments solved problem of symbolic regression of dynamic system. The number of iterations needed for required quality of regression was used as the measure of (p)RNG influence. These experiments point that different (p)RNGs fit to different evolutionary operators, that some combinations (p)RNGs are better than others and that some theoretically excellent (p)RNGs produces poor results. Presented experiments point that the efficiency of evolutionary algorithms might be increased by application of more (p)RNGs in one algorithm optimised for each particular evolutionary operator.

  • [6] Contents volume 27 (2017), ... 607
  • [7] Authors index volume 27 (2017), ...611

5/2017

  • [1] Goltsev A., Gritsenko V., Húsek D. (UA, CZ) ,
    Extraction of homogeneous fine-grained texture segments in visual images, 447-477

      Full text     DOI: http://dx.doi.org/10.14311/NNW.2017.27.024

    Abstract: A new heuristic algorithm is proposed for extraction of all homogeneous fine-grained texture segments present in any visual image. The segments extracted by this algorithm should comply with human understanding of homogeneous fine-grained areas. The algorithm sequentially extracts segments from more homogeneous to less homogeneous ones. The algorithm belongs to a region growing approach. So, for each segment, an initial seed point of this segment is found. Then, from this initial pixel, the segment begins to expand occupying its adjacent neighborhoods. This procedure of expansion of the segment continues till the segment reaches its borders. The algorithm examines neighboring pixels using texture features extracted in the image by means of a set of texture windows. The segmentation process terminates when the image contains no more sizable homogeneous segments. The segmentation procedure is fully unsupervised, i.e., it does not use a priori knowledge on either the type of textures or the number of texture segments in the image. Using black and white natural scenes, a series of experiments demonstrates efficiency of the algorithm in extraction of homogeneous fine-grained texture segments and the segmentation looks reasonable "from a human point of view".

  • [2] Provinský P. (CZ)
    Floppy logic – a younger sister of fuzzy logic, 479-497

      Full text     DOI: http://dx.doi.org/10.14311/NNW.2017.27.025

    Abstract: This article introduces a floppy logic - a new method of work with fuzzy sets. This theory is a nice connection between the logic, the probability theory and the fuzzy sets. The floppy logic has several advantages compared to the fuzzy logic: All propositions, which are equivalent in the bivalent logic, are equivalent in the floppy logic too. Logical operations are modeled unambiguously, not by using many alternative t-norms and t-conorms. In floppy logic, we can use the whole apparatus of Kolmogorov's probability theory. This theory allows to work consistently with systems that are described by fuzzy sets, probability distributions and accurate values simultaneously.

  • [3] Zhou Z., Chen J., Song Y., Zhu Z., Liu X. (China)
    RFSEN-ELM: Selective ensemble of extreme learning machines using rotation forest for image classification, 499-517

      Full text     DOI: http://dx.doi.org/10.14311/NNW.2017.27.026

    Abstract: Learning machine (ELM), as a new learning mechanism for single hidden layer feedforward neural networks (SLFNs), has shown its advantages, such as fast computation speed and good generalization performance. However, the weak robustness of ELM is an unavoidable defect for image classification. To address the problem, we propose a novel ensemble method which combines rotation forest and selective ensemble model in this paper. Firstly, ELM and rotation forest are integrated to construct an ensemble classifier (RF-ELM), which combines the advantages of both rotation forest and ELM. The purpose of rotation forest here is to enhance the diversity of each base classifier which can improve the performance generalization. Then several ELMs are removed from the ensemble pool by using genetic algorithm (GA) based selective ensemble model to further enhance the robustness.Finally, the remaining ELMs are grouped as a selected ensemble classifier (RFSEN-ELM) for image classification. The performance is analyzed and compared with several existing methods on benchmark datasets and the experimental results demonstrate that the proposed algorithm substantially improves the accuracy and robustness of classification at an acceptable level of training cost.

  • [4] Temel T. (Turkey)
    A single-step clustering algorithm based on a new information-theoretic sample association metric definition, 519-528

      Full text     DOI: http://dx.doi.org/10.14311/NNW.2017.27.027

    Abstract: A single-step information-theoretic algorithm that is able to identify possible clusters in dataset is presented. The proposed algorithm consists in representation of data scatter in terms of similarity-based data point entropy and probability descriptions. By using these quantities, an information-theoretic association metric called mutual ambiguity between data points is defined, which then is to be employed in determining particular data points called cluster identifiers. For forming individual clusters corresponding to cluster identifiers determined as such, a cluster relevance rule is defined. Since cluster identifiers and associative cluster member data points can be identified without recursive or iterative search, the algorithm is single-step. The algorithm is tested and justified with experiments by using synthetic and anonymous real datasets. Simulation results demonstrate that the proposed algorithm also exhibits more reliable performance in statistical sense compared to major algorithms.

  • [5] Sarma H., Samaddar A.B., Porzel R., Smeddinck J.D., Malaka R. (DE, India) ,
    Updating Bayesian networks using crowds, 529-540

      Full text     DOI: http://dx.doi.org/10.14311/NNW.2017.27.028

    Abstract: Physical exercise instruction sheets are often difficult to understand. In most cases considerable information remains implicit, which also poses a considerable hurdle regarding potential machine understanding. Major missing information types include the source and destination location of a human movement. We present a Bayesian network for extracting the implicit or missing information from typical exercise instruction sheets. We propose two different kinds of Bayesian networks which consist of three and four variables respectively. The network with three variables is designed for single exercise instruction, featuring a single action or pose. The other is designed for single of multiple sentence(s) spanning two actions or poses. The conditional probability table (CPT) is the backbone of a Bayesian network. At the start, the CPT is informed by our physical exercise instruction sheet corpus (PEISC). Keeping the Action and Bodypart fixed, we have developed our CPT using a crowdsourcing. We have developed a CPT update system using 13 different exercises consisting of 44 different exercise videos. Using this system, candidate exercise executions are rated by participants and their ratings update the CPT. We also update the Action variable, which consists of 14 different values (action verbs) using crowdsourcing with a human computation approach.


4/2017

  • [1] Lom M., Pribyl O. (CZ)
    Modeling of Smart City Building Blocks Using Multi-Agent Systems, 317-331

      Full text     DOI: http://dx.doi.org/10.14311/NNW.2017.27.018

    Abstract: Technology has undergone rapid development in the past several decades and we are now at a point where many technologies are available to help create smart cities. Many technology companies and research institutions as well as political organizations are currently discussing this field with the highest priority. One can say that the biggest challenge to smart cities is not technologies themselves, but the merging of all available technologies into one symbiotic unit that fulfills the expected objectives. Smart cities are about connecting subsystems, sharing and evaluating data, and providing quality of life and satisfaction to citizens. We have various models of transportation systems, optimizations of energy usage, street lighting systems, building management systems, urban transport optimizations, however currently, such models are dealt with separately. In this paper, we provide an overview of the smart city concept and discuss why Multi-agent systems are the right tool for the modeling of smart cities. The biggest challenge is in connecting and linking particular subsystems within a smart city. In this paper, a modeling of a smart city building blocks is provided and demonstrated with one particular example -- a smart street lighting system. Focus will be on the decomposition of the system into subsystems as well as a description of particular modules. We propose to build models and since each individual entity can be modeled as an agent with its beliefs, desires and intentions, we suggest using Multi-agent systems as a tool for modeling systems` connections within the smart city and assessing how best to use the data from those systems.

  • [2] Rulc V., Purš H., Kovanda J. (CZ)
    Analysis of controlled mechanism with significant nonlinearities, 333-349

      Full text     DOI: http://dx.doi.org/10.14311/NNW.2017.27.019

    Abstract: Solving inverted pendulum by co-simulation between multi-body solver MotionSolve and signal processing control in solidThinking Activate. The simulation of inverted pendulum uses an innovative model of friction which is physically and mathematically more accurate than usual CAE friction models. This model of friction adds nonlinearity to the system. Two types of controlling mechanism for active balancing of inverted pendulum are used: PID and ANN controller. A non-traditional false angular deviation approach for returning a cart to its initial position was used.

  • [3] Dolezel P., Heckenbergerova J. (CZ)
    Computationally Simple Neural Network Approach to Determine Piecewise-Linear Dynamical Model, 351-371

      Full text     DOI: http://dx.doi.org/10.14311/NNW.2017.27.020

    Abstract: The article introduces a new technique for nonlinear system modeling. This approach, in comparison to its alternatives, is straight and computationally undemanding. The article employs the fact that once a nonlinear problem is modeled by a piecewise-linear model, it can be solved by many efficient techniques. Thus, the result of introduced technique provides a set of linear equations. Each of the equations is valid in some region of state space and together, they approximate the whole nonlinear problem. The technique is comprehensively described and its advantages are demonstrated on an example.

  • [4] Hozman J., Bradác J., Kovanda J. (CZ)
    DG solver for the simulation of simplified elastic waves in two-dimensional piecewise homogeneous media, 373-389

      Full text     DOI: http://dx.doi.org/10.14311/NNW.2017.27.021

    Abstract: The theory of elasticity is a very important discipline which has a lot of applications in science and engineering. In this paper we are interested in elastic materials with different properties between interfaces implicated the discontinuous coefficients in the governing elasticity equations. The main aim is to develop a practical numerical scheme for modeling the behaviour of a simplified piecewise homogeneous medium subjected to an external action in 2D domains. Therefore, the discontinuous Galerkin method is used for the simulation of elastic waves in such elastic materials. The special attention is also paid to treatment of boundary and interface conditions. For the treatment of the time dependency the implicit Euler method is employed. Moreover, the limiting procedure is incorporated in the resulting numerical scheme in order to overcome nonphysical spurious overshoots and undershoots in the vicinity of discontinuities in discrete solutions. Finally, we present computational results for two-component material, representing a planar elastic body subjected to a mechanical hit or mechanical loading.

  • [5] Theuer M. et al. (CZ)
    Efficient methods of automatic calibration for rainfall-runoff modelling in the Floreon+ system, 391-414

      Full text     DOI: http://dx.doi.org/10.14311/NNW.2017.27.022

    Abstract: Calibration of rainfall-runoff model parameters is an inseparable part of hydrological simulations. To achieve more accurate results of these simulations, it is necessary to implement an efficient calibration method that provides sufficient refinement of the model parameters in a reasonable time frame. In order to perform the calibration repeatedly for large amount of data and provide results of calibrated model simulations for the flood warning process in a short time, the method also has to be automated. In this paper, several local and global optimization methods are tested for their efficiency. The main goal is to identify the most accurate method for the calibration process that provides accurate results in an operational time frame (typically less than 1 hour) to be used in the flood prediction Floreon\textsuperscript{+} system. All calibrations were performed on the measured data during the rainfall events in 2010 in the Moravian-Silesian region (Czech Republic) using our in-house rainfall-runoff model.

  • [6] Garlík B. (CZ)
    The application of artificial intelligence in the process of optimizing energy consumption in intelligent areas, 415-446

      Full text     DOI: http://dx.doi.org/10.14311/NNW.2017.27.023

    Abstract: The text describes the optimization task of renewable energy sources distributed to electrical microgrid of fictitious intelligent area that consists of intelligent buildings. Firstly, to solve this task a general optimization heuristic method of simulated annealing will be described. Testing was performed on the analytical functions but those will be only covered marginally. Of the tests on the approximation functions the method of simulated annealing would be the most suitable algorithm for the optimization task. Furthermore, two experiments were introduced. The first lies in the application of cluster analysis on daily diagrams of electricity consumption in intelligent buildings. Because the modeled year history of hourly electricity consumption is represented by multidimensional data this data forms the training set during the adaptive dynamics submitted to a competence model of neural network by days. After the network adaptation process the Kohonen's map during the adaptive dynamics will be drawn, from which required clusters can be read. In the second experiment a sorting design of the resources for typical days of a week is performed in the computer program UniCon.


3/2017

  • [1] D. Balara, J. Timko, J. Zilková, M. Leso (SK)
    Neural networks application for mechanical parameters identification of asynchronous motor, 259-270

      Full text     DOI: http://dx.doi.org/10.14311/NNW.2017.27.013

    Abstract: A method for identification of mechanical parameters of an asynchronous motor is presented in this paper. The identification method is based on the use of our knowledge of the system. This paper clarifies the method by using the example identifying of mechanical parameters of the three-phase squirrel-cage asynchronous motor.A model of mechanical subsystem of the motor is presented as well as results of simulation. The special neural network is used as an identification model and its adaptation is based on the gradient descent method.The parameters of mechanical subsystem are derived from the values of synaptic weights of the neural identification model after its adaptation. Deviation of identified mechanical parameters in the case of moment inertia was up to 0.03% and in the case of load torque was 1.45% of real values.

  • [2] D. Xu, L. Zhang, H. Zhang A. (China)
    Learning Algorithms in Quaternion Neural Networks Using GHR Calculus , 271-282

      Full text     DOI: http://dx.doi.org/10.14311/NNW.2017.27.014

    Abstract: One difficulty for quaternion neural networks (QNNs) is that quaternion nonlinear activation functions are usually non-analytic and thus quaternion derivatives cannot be used. In this paper, we derive the quaternion gradient descent, approximated quaternion Gauss-Newton and quaternion Levenberg-Marquardt algorithms for feedforward QNNs based on the GHR calculus, which is suitable for analytic and non-analytic quaternion functions. Meanwhile, we solve a widely linear quaternion least squares problem in the derivation of quaternion Gauss-Newton algorithm, which is more general than the usual least squares prob¬lem. A rigorous analysis of the convergence of the proposed algorithms is provided. Simulations on the prediction of benchmark signals support the approach.

  • [3] M.S.A Siddiquee, S. Hoque. (Bangladesh)
    Predicting the daily traffc volume from hourly traffc data using artificial neural network, 283-294

      Full text     DOI: http://dx.doi.org/10.14311/NNW.2017.27.015

    Abstract: The prediction of traffic volume over time is very important to control the flow of traffic on a road network. Traffic count is usually averaged over time to predict for the larger time domain. This paper aims at finding the detail variation of a systematic survey of hourly traffic volume data over a time of four years along the North Bengal corridor of Bangladesh (at Jamuna toll collection point) and its equivalent numerical model by using a Artificial Neural Network. The Neural Network is trained with the intermittent data of 13 weeks over four years and the missing data is interpreted with quite reasonable accuracy (12.67% MAE) with this ANN model. The ANN model captured the variety of trends of the traffic data very accurately as has been depicted in the paper

  • [4] Muhammad Waqas, Abdul Aziz Bhatti (Pakistan)
    Optimization of N+1 Queens Problem using Discrete Neural Network, 295-308

      Full text     DOI: http://dx.doi.org/10.14311/NNW.2017.27.016

    Abstract: Combinatorial optimization problems are extensively solved by using neural networks. Hopfield-Tank model is used to solve Traveling Salesman Problem and many NP-Hard Problems. This paper describes a neural network optimizer/scheduler that optimizes a solution for a highly complicated version of N Queens Problem (NQP), i.e. N+1 non-threatening Queens on a N*N chessboard with an intermediate pawn on it. Both synchronous and asynchronous methods of updating of the neurons have been applied for optimization of N+1 Queens Problem. Computer simulations are used to confirm the results. The proposed neural network is attracted to optimized solution or finds the global minima in 90% of the trials. A new rule of initialization, i.e. the proximity rule of initialization has been proposed. Using the proximity rule of initialization the performance of the system is enhanced and the system converges to an optimal solution in much less time. Many novel applications like multiprocessor job scheduling, resource optimization, of the above mentioned algorithm have been proposed. N Queens Problem has been solved by many techniques but no other algorithm exists to solve N+1 QP in the literature. Consequently, the performance of the network is compared with full space search algorithm.

  • [5] Kilic E., Dundar P. (Turkey)
    Total accessibility number of graphs, 309-315

      Full text     DOI: http://dx.doi.org/10.14311/NNW.2017.27.017

    Abstract: One of the most important problems in communication network design is the stability of network after any disruption of stations or links. Since a network can be modeled by a graph, this concept is examined under the view of vulnerability of graphs. There are many vulnerability measures that were defined in this sense. In recent years, measures have been defined over some vertices or edges having specific properties. These measures can be considered to be a second type of measures. Here we define a new measure of the second type called the total accessibility. This measure is based on accessible sets of a graph. In our study we give the total accessibility number of well known graph models such as Pn, Cn, Km,n, W1,n, K1,n. We also examine this new measure under operations on graphs. A simple algorithm, which calculates the total accessibility number of graphs, is given. We observe that when any two graphs of the same size are compared in stability, it is inferred that the graph of higher total accessibility number is more stable than the other one. All the graphs considered in this paper are undirected, loopless and connected.


2/2017

  • [1] Kuklová J., Přibyl O. (CZ)
    Changeover from decision tree approach to fuzzy logic approach within highway management , 181-196

      Full text     DOI: http://dx.doi.org/10.14311/NNW.2017.27.008

    Abstract: {abstract} This paper deals with the changeover from the decision tree (bivalent logic) approach to the fuzzy logic approach to highway traffic control, particularly to variable speed limit displays. The usage of existing knowledge from decision tree control is one of the most suitable methods for identification of the new fuzzy model. However, such method introduces several difficulties. These difficulties are described and possible measures are proposed. Several fuzzy logic algorithms were developed and tested by a~microsimulation model. The results are presented and the finest algorithm is recommended for testing on the Prague City Ring Road in real conditions. This paper provides a~guidance for researchers and practitioners dealing with similar problem formulation.

  • [2] Popa M.C., Rothkrantz L.J.M., Wiggers P., Shan C. (NL)
    Assessment of Facial Expressions in Product Appreciation, 197-214

      Full text     DOI: http://dx.doi.org/10.14311/NNW.2017.27.009

    Abstract: In the marketing area, new trends are emerging, as customers are not only interested in the quality of the products or delivered services, but also in a stimulating shopping experience. Creating and influencing customers' experiences has become a valuable differentiation strategy for retailers. Therefore, understanding and assessing the customers' emotional response in relation to products/services represents an important asset. The purpose of this paper consists of investigating whether the customer's facial expressions shown during product appreciation are positive or negative and also which types of emotions are related to product appreciation. We collected a database of emotional facial expressions, by presenting a set of forty product related pictures to a number of test subjects. Next, we analysed the obtained facial expressions, by extracting both geometric and appearance features. Furthermore, we modeled them both in an unsupervised and supervised manner. Clustering techniques proved to be efficient at differentiating between positive and negative facial expressions in 78\% of the cases. Next, we performed more refined analysis of the different types of emotions, by employing different classification methods and we achieved 84\% accuracy for seven emotional classes and 95\% for the positive vs. negative.

  • [3] Seidlová R., Pozivil J., Seidl J., Malecl L. (CZ)
    Synthetic data generator for testing of classification rule algorithms, 215-229

      Full text     DOI: http://dx.doi.org/10.14311/NNW.2017.27.010

    Abstract: We developed a data generating system that is able to create systematically testing datasets that accomplish user’s requirements such as number of rows, number and type of attributes, number of missing values, class noise and imbalance ratio. These datasets can be used for testing of the algorithms designed for solving classification rule problem. We used them for optimizing of the parameters of the classification algorithm based on the behavior of ant colonies. But they can be advantageously used for other applications too. Program generates output files in ARFF format. Two standards and one user-define probability distributions are used in data generation: uniform distribution, normal distribution and irregular distribution for nominal attributes. To our knowledge, our system is probably the first synthetic data generation system that systematically generates datasets for examination and judgment of the classification rule algorithms.

  • [4] Gadri S., Moussaoui A. (Algeria)
    Application of a New Set of Pseudo-Distances in Documents Categorization, 231-245

      Full text     DOI: http://dx.doi.org/10.14311/NNW.2017.27.011

    Abstract: Automatic text classification is a very important task that consists in assigning labels (categories, groups, classes) to a given text based on a set of previously labeled texts called training set. The work presented in this paper treats the problem of automatic topical text categorization. It is a supervised classification because it works on a predefined set of classes and topical because it uses topics or subjects of texts as classes. In this context, we used a new approach based on $k$-NN algorithm, as well as a new set of pseudo-distances (distance metrics) known in the field of language identification. We also proposed a simple and effective method to improve the quality of performed categorization.

  • [5] Ye W., Liu S., Liu X. (China)
    Transition modes between spiking and bursting in a pacemaker neuron, 247-258

      Full text     DOI: http://dx.doi.org/10.14311/NNW.2017.27.012

    Abstract: Central pattern generators (CPGs) play an important role in controlling rhythmic movements in vivo. Increased insight into mechanisms of CPGs can be obtained by perturbing neuron activities so as to study a range of behaviors. By applying this method, a series of simulations were performed to research different transition modes between firing patterns in a pacemaker neuron model of stomatogastric ganglion (STG). Firstly, with the perturbation of parameters in model, such as external stimulus, parameters in compartments and connection between compartments, different firing patterns and bifurcation of inter-spike intervals (ISIs) were obtained to exhibit the impact of single parameter on the transions between spiking and bursting. Moreover, perturbing two parameters gCa, Iext simultaneously induced the continuous variation of the bifurcation mode, which implied the crucial role of calcium channel in regulating the rhythm generation. Finally, a two-dimensional parameter space (gCa, Iext) was constructed by spike-counting method to capture the distribution of the firing patterns and different transition mode between them in a comprehensive aspect. In this parameter space, three basic transition modes were concluded: bifurcation ring, period-doubling mode and period-adding mode.


1/2017

  • Editorial Board of Neural Network World (CZ)
    Petr Hájek passed away, 1-2,       
    Full text



  • [1] Chaudhary P., Gupta P.P. (India)
    A Novel Framework to Alleviate Dissemination of XSS Worms in Online Social Network (OSN) using View Segregation , 5-26

      Full text     DOI: http://dx.doi.org/10.14311/NNW.2017.27.001

    Abstract: In this paper, we propose a client-server based framework that alleviates the dissemination of XSS worms from the OSN. The framework initially creates the views corresponding to retrieved request on the server-side. Such views indicate that which part of the generated web page on the server can be accessed by user depending on the generated Access Control List (ACL). Secondly, JavaScript attack vectors are retrieved from the HTTP response by referring the blacklist repository of attack vectors. Finally, injection of sanitization primitives will be done on the client-side in place of extracted JavaScript attack vectors. The framework will perform the sanitization on such attack vectors strictly in a context-aware manner. The experimental testing of our framework has performed on the two platforms of open source OSN-based web applications. The observed detection rate of JavaScript attack vectors was effective and acceptable as compared to other existing XSS defensive methodologies. The proposed framework has optimized the method of auto-context-aware sanitization in contrast to other existing approaches and hence incurs a low and acceptable performance overhead.

  • [2] S. Kamal, N. Dey, A. S. Ashour, S. Ripon,V. E. Balas, M. S. Kaysar (Bangladesh)
    FbMapping: An Automated System for Monitoring Facebook Data , 27-58

      Full text     DOI: http://dx.doi.org/10.14311/NNW.2017.27.002

    Abstract: In recent modernized era, the number of the Facebook users is increasing dramatically. Moreover, the daily life information on social networking sites is changing energetically over web. Teenagers and university students are the major users for the different social networks all over the world. In order to maintain rapid user satisfactions, information flow and clustering are essential. However, these tasks are very challenging due to the excessive datasets. In this context, cleaning the original data is significant. Thus, in the current work the Fishers Discrimination Criterion (FDC) is applied to clean the raw datasets. The FDC separates the datasets for superior fit under least square sense. It arranges datasets by combining linearly with greater ratios of between -- groups and within the groups. In the proposed approach, the separated data are handled by the Bigtable mapping that is constructed with Map specification, tabular representation and aggregation. The first phase organizes the cleaned datasets in row, column and timestamps. In the tabular representation, Sorted String Table (SSTable) ensures the exact mapping. Aggregation phase is employed to find out the similarity among the extracted datasets. Mapping, preprocessing and aggregation help to monitor information flow and communication over Facebook. For smooth and continuous monitoring, the Dynamic Source Monitoring (DSM) scheme is applied. Adequate experimental comparisons and synthesis are performed with mapping the Facebook datasets. The results prove the efficiency of the proposed machine learning approaches for the Facebook datasets monitoring.

  • [3] Yang Z., Li Z., Fan K., Huang J. (China)
    Exploiting Multi-Sources Query Expansion in Microblogging Filtering , 59-76

      Full text     DOI: http://dx.doi.org/10.14311/NNW.2017.27.003

    Abstract: Microblogging filtering is intended to filter out irrelevant content, and select useful, new, and timely content from microblogs. However, microblogging filtering suffers from the problem of insufficient samples which renders the probabilistic models unreliable. To mitigate this problem, a novel method is proposed in this study. It is believed that an explicit brief query is only an abstract of the user's information needs, and it’s difficult to infer users' actual searching intents and interests. Based on this belief, a filtering model is built where the multi-sources query expansion in microblogging filtering is exploited and expanded query is submitted as user’s interest. To manage the external expansion risk, a user filter graph inference method is proposed, which is characterized by combination of external multi-sources information, and a risk minimization filtering model is introduced to achieve the best reasoning through the multi-sources expansion. A series of experiments are conducted to evaluate the effectiveness of proposed framework on an annotated tweets corpus. The results of these experiments show that our method is effective in tweets retrieval as compared with the baseline standards.

  • [4] Yuan W., Guan D. (China)
    Optimized Trust-Aware Recommender System using Genetic Algorithm, 77-94

      Full text     DOI: http://dx.doi.org/10.14311/NNW.2017.27.004

    Abstract: Trust-aware recommender system (TARS) recommends ratings based on user trust. It greatly improves the conventional collaborative filtering by providing reliable recommendations when dealing with the data sparseness problem. One basic research issue of TARS is to improve the recommending efficiency, in which the key point is to find sufficient number of recommenders efficiently for active users. Existing works searched recommenders via a skeleton, which consists of a number of hub nodes. The hub nodes are those who have superior degrees based on the scale-freeness of the trust network. However, existing works did not consider the skeleton maintenance cost and the coverage overlap between nodes of the skeleton. They also failed to suggest the proper size of the skeleton. This paper proposes an optimized TARS model to solve the problems of existing works. By using the genetic algorithm, our model chooses the most suitable nodes for the skeleton of recommender searching. It can achieve the maximum prediction coverage with the minimum skeleton maintenance cost. Simulation results show that compared with existing works, our model can reduce more than 90{\%} of the skeleton maintenance cost with reasonable prediction coverage.

  • [5] H. Li, T. Zhao, N. Li, Q. Cai, J. Du (China)
    Feature Matching of Multi-view 3D Models Based on Hash Binary Encoding, 95-106

      Full text     DOI: http://dx.doi.org/10.14311/NNW.2017.27.005

    Abstract: Image data and 3D model data have emerged as resourceful foundation for information with proliferation of image capturing devices and social media. In this paper, a feature matching method based on hash binary encoding for multi-view 3D models in social media is proposed. SIFT algorithm is first used to extract features of the depth image, and then RANSAC is utilized as a filter. Finally, a cascade hash binary encoding algorithm is adapted to match the feature of multi-view models. Experimental results on SHREC2014 dataset have shown the effectiveness of the proposed method.

  • [6] A.A. Frolov, D. Húsek, E.V. Biryukova, P.D. Bobrov, O.A. Mokienkoy, A.V. Alexandrov (Russia,CZ) ,
    Principles of motor recovery in post-stroke Patients Using Hand Exoskeleton Controlled by the Brain-Computer Interface Based on Motor Imagery , 107-138

      Full text     DOI: http://dx.doi.org/10.14311/NNW.2017.27.006

    Abstract: Motor recovery in post-stroke and post-traumatic patients using exoskeleton controlled by the brain-computer interface (BCI) is a new and promising rehabilitation procedure. Its development is a multidisciplinary research which requires, the teamwork of experts in neurology, neurophysiology, physics, mathematics, biomechanics and robotics. Some aspects of all these fields of study concerning the development of this rehabilitation procedure are described in the paper. The description includes the principles and physiological prerequisites of BCI based on motor imagery, biologically adequate principles of exoskeleton design and control and the results of clinical application.

  • [7] T.M. Usha, S. Appavu alias Balamurugan (India)
    Computational Modeling of Electricity Consumption Using Econometric Variables Based on Neural Network Training Algorithms, 139-178

      Full text     DOI: http://dx.doi.org/10.14311/NNW.2017.27.007

    Abstract: Recently, there has been a significant emphasis in the forecasting of the electricity demand due to the increase in the power consumption. Energy demand forecasting is a very important task in the electric power distribution system to enable appropriate planning for future power generation. Quantitative and qualitative methods have been utilizedpreviously for the electricity demand forecasting. Due to the limitations inthe availability of data, these methods fail to provide effective results. With the development of the advanced tools, these methods are replaced by efficient forecasting techniques. This paper presents the computational modeling of electricity consumption based on the Neural Network (NN) training algorithms. The main aim of the work is to determine the optimal training algorithm for electricity demand forecasting. From the experimental analysis, it is concluded that the Bayesian regularization training algorithm exhibits low relative error and high correlation coefficient than other training algorithms. Thus, the Bayesian Regularization training algorithm is selected as the optimal training algorithm for the effective prediction of the electricity demand. Finally, the economic input attributes are forecasted for next 15 years using time series forecasting. Using this forecasted economic attributes and with the optimal Bayesian Regularization training algorithm, the electricity demand for the next 15 years ispredicted. The comparative analysis of the NN training algorithms for the proposed dataset and larger datasets obtained from the UCI repository and American Statistical Association shows that the Bayesian Regularization training algorithm yields higher correlation value and lower relative error than other training algorithms.