Contents of Volume 34 (2024)

1/2024

2/2024

  • [1] Yi D., Kim I., Bu S., (Korea)
    Variations of Training Process in Vanilla Recurrent Neural Network Framework, pp. 73-87

      Full text         https://doi.org/10.14311/NNW.2024.34.005

    Abstract: Recurrent neural networks (RNNs) are capable of learning features and long term dependencies from sequential and time series data and show outstanding performance in sequential modeling tasks. However, training process in RNNs is troubled by issues in learning processes such as slow inference, vanishing gradients and difficulties in capturing long term dependencies. In this paper, we introduce a new learning technique to update the weight set as we change the input sequence which is shifted by certain amount of time in training process, instead of using a traditional way to calculate one set of the weights and bias in training time series with sequences shifted by certain amount of time series. We also consider an algorithm for an evaluation process. In the traditional way, the evaluation process is executed by using final weights and biases calculated in the training process. Instead, during the testing process, the weights and biases are iteratively updated in each sequence as done in the training process. Several numerical experiments demonstrate the efficiency of the proposed techniques.

  • [2] Uglickich E., Nagy I , (CZ)
    3D Local Crime Type Models Based on Crime Hotspot Detection, pp. 89-110

      Full text         https://doi.org/10.14311/NNW.2024.34.006

    Abstract: This paper deals with the analysis of the relationship between locations and types of crime observed in the Czech Republic. Cluster analysis of crime data based on the recursive Bayesian mixture estimation algorithm is used to identify crime hotspots and estimate local models of crime type. The experiments report that the 2D configuration of the algorithm allows the detection of crime hotspots online. The 3D configuration provides 29% more accurate crime type models than 2D clustering and alternative data mining algorithms. For the data set used, it was determined in which crime hotspots the most serious and most frequent types of crime can be expected to occur with the highest probability. The limitation of the study is the artificial support of the 3D clusters by the fully continuous data vector with the recoded values of the crime type. The potential use of the algorithm is expected in online web applications for sharing information on criminal offenses managed by the Police of the Czech Republic with the public and local government entities in the Czech Republic.

  • [3] Kovanda J., Hozman J. , (CZ)
    Equations of Motion of Mechanical Systems with Switchable Constraints, pp. 111-134

      Full text         https://doi.org/10.14311/NNW.2024.34.007

    Abstract: The article deals with a numerical and analytical approach to solving the equations of motion, which enables to treat the considered problems with the change of system structure or number of degrees of freedom without interrupting the numerical integration process. The described methodology allows effectively incorporate switchable constraints in the systems in accordance with their flexible structures. The crucial idea is based on the formulation of the resulting differentialalgebraic equations into a saddle point system, where the switchable constraints are represented by a sign matrix with variable rank. In connection with this property, a pseudoinversion is applied to eliminate algebraic variables and transform the problem to the first order system of ordinary differential equations. Moreover, the time independent case leads to linear autonomous systems with non-diagonalizable matrices, as is proved. The relevant numerical scheme is based on Runge-Kutta methods, that correspond to the power series of the resulting matrix exponential for time independent problems. The methodology presented is illustrated on the idealized two-mass oscillator with a switchable constraint. The numerical experiments performed range from initial stages, through simple transient cases to damped intentional control. The advanced applications can be found in robotics, active and controlled systems, and in the simulations of complex systems in biology and related areas. Moreover, the methodology can also be applied in the simulation of transport systems, especially in relation to vehicle technology, a quarter car suspension system, a vibration control mechanism, a torsion system with a clutch, and machine balancing and storage should to be highlighted.


1/2024

  • [1] Sudharson D., Gomathi R., Selvam L., (India)
    Software Reliability Analysis by Using the Bidirectional Attention Based Zeiler-Fergus Convolutional Neural Network, pp. 1-25

      Full text         https://doi.org/10.14311/NNW.2024.34.001

    Abstract: Software quality assurance relies heavily on software reliability as one of its primary metrics. Numerous studies have been conducted to identify the software reliability. Improved software dependability may be studied using a triangular approach that includes software modeling, measurement, and improvement. Each of these steps is critical to the development of a solid software system. Improved accuracy in calculating dependability is critical to managing the quality of software. It has been discovered that deep learning algorithms are excellent methods of assessing many aspects of software dependability. Software systems contain distinct characteristics that can be addressed using deep learning techniques. In this study, a deep-learning-based bidirectional attention-based Zeiler-Fergus convolutional neural network (BA-ZFCNN) technique has been suggested to assess software dependability. In the beginning, the data were standardized by using the scalable error splash method. This approach was then used to extract the software fault-related characteristics using hypertuned evolutionary salp swarm optimization (HESSO). Finally, the Zeiler-Fergus convolutional neural network based on bidirectional attention (BA-ZFCNN) may be used to assess software dependability. The suggested method is used to forecast how many defects or failures there are in a software product. AR1 software defect data is widely used to test the effectiveness of deep learning and traditional machine learning methods. The experimental results reveal that the proposed method’s accuracy (96.7%) is higher than the current techniques’ accuracy.

  • [2] Beneš V., Svítek M., Michalíková A., Melicherčík M. , (CZ, SK) ,
    Situation Model of the Transport, Transport Emissions and Meteorological Conditions, pp. 27-36

      Full text         https://doi.org/10.14311/NNW.2024.34.002

    Abstract: Air pollution in cities and the possibilities of reducing this pollution represent one of the most important factors that today’s society has to deal with. This paper focuses on a systemic approach to traffic emissions with their relation to meteorological conditions, analyzing the effect of weather on the quantity and dispersion of traffic emissions in a city. Using fuzzy inference systems (FIS) the model for predicting changes in emissions depending on various conditions is developed. The proposed model is based on traffic, meteorology and emission data measured in Prague, Czech Republic. The main objective of the work is to provide insight into how urban planners and policymakers can plan and manage urban transportation more efficiently with environmental protection in mind.

  • [3] Halim A.A., Mustafa W.A., Nasir A.S.A., Ismail S., Alquran H., (Malaysia, Jordan) ,
    Nucleus Cell Segmentation on Pap Smear Image Using Bradley Modification Algorithm, pp. 37-51

      Full text         https://doi.org/10.14311/NNW.2024.34.003

    Abstract: Early detection of cervical cancer can help patients obtain the best treatment through various means. In general, computer-aided diagnosis has a high impact on the accuracy, reliability, and convenience of cervical cancer. However, several limitations have been faced through the design process in detecting or classifying the cells, such as variation of image features and low-image resolution. Moreover, shape indifference is one of the limitations in terms of image processing scope. The metrics used to measure the size and shape of the cells have not been developed to distinguish the differences between the shape of the objects. This paper focused on the detection and segmentation of the nucleus cell region in Pap smear images based on Bradley local thresholding. The proposed method evolved several steps, such as color adjustment, k-means, and a Bradley modification algorithm. Based on image quality assessment (IQA), the numerical evaluation results indicate that the proposed approach has segmented a full area of the nucleus cell region significantly and efficiently compared to the original Bradley algorithm. We obtained F-measure (98.62%), sensitivity (99.13%), and accuracy (97.96%). It has also been proven that the proposed method can effectively address the issue of low contrast and black noise. Hence, the proposed method differs from the previous research in terms of color disproportion adjustment and the modification of Bradley’s algorithm for Pap smear image convenience.

  • [4] Saranya K., Pandiyan P.M., Hema C.R. , (India)
    Certain Investigations on Feature Selection Technique Using Artificial Immune Systems for EEG Color Visualization Classification, pp. 53-71

      Full text         https://doi.org/10.14311/NNW.2024.34.004

    Abstract: This paper aims to extract and select the significant features of electroencephalogram (EEG) signals and classify the visual stimulation of distinct colors. In this work, a novel method for selecting distinct colors using EEG signals called affinity artificial immune and Daubechies wavelet time-based learning (AAIDWTL) is proposed. Initially, the EEG signals were collected in a controlled environment and an in-built band-pass filter was applied to remove the artifacts. The filtered signals were converted into frequency domain signals using least squarebased short-term Fourier transform. After that, by utilizing Daubechies wavelet statistical time-based feature extraction model the time domain features were extracted. Followed by, computationally efficient features were selected using an affinity artificial immune-based feature selection model. The selected features were classified using a polynomial kernel multiclass classification-based machine learning algorithm and achieved an accuracy of 97.5% when compared with other methods like linear discriminant analysis (LDA) which obtained only 92%. Furthermore, while utilizing the proposed method classification time was considerably less when compared to LDA. The experimental result shows that the proposed color stimulation of the EEG signals method achieved greater improvement in terms of both classification time and classification accuracy with a minimum false positive rate.