Vol. 17 No. 2 (2023)
Articles
Abstract: Speech emotion recognition (SER) is a challenging task in the field of artificial intelligence and machine learning. Over the years, researchers have proposed various approaches to recognize emotions from speech signals. This article will analyze and discuss some of the previous works on machine learning and deep learning in SER. This survey study focuses on the importance of the human voice in determining emotional and psychological states. Various methods were used to successfully classify emotions such as anger, sadness, happiness, fear, disgust, neutral, and surprise. Reviewed in this study was conducted in sequential stages including pre-processing treatment, feature selection, classification, and evaluation of results. different data sets were also reviewed for international languages such as English, Hindi, German, Urdu, Tamil, French, and Arabic. The study primarily focused on artificial intelligence and machine learning algorithms due to their flexibility and ease of understanding with distinct results.
Abstract: In this paper, we used the modified Adomian decomposition(MADM) method to solve integral equations of the Volterra and Fredholm type as well, and then the Adomian method was combined with the bee algorithm(ABC) and obtained values for the parameter that improved the results obtained by solving some examples and were more accurate than the default method, These results are illustrated by calculating the maximum absolute errors (MAE) and mean squared errors (MSE).
Abstract: Recently, the increase in the emergence of fake videos that have a high degree of accuracy makes it difficult to distinguish from real ones. This is due to the rapid development of deep-learning techniques, especially Generative Adversarial Networks (GAN). The harmful nature of deepfakes urges immediate action to improve the detection of such videos. In this work, we proposed a new model to detect deepfakes based on a hybrid approach for feature extraction by using 128-identity features obtained from facenet_CNN combined with most powerful 10-PCA features. All these features are extracted from cropped faces of 10 frames for each video. FaceForensics++ (FF++) dataset was used to train and test the model, which gave a maximum test accuracy of 0.83, precision of 0.824 and recall value of 0.849.
Abstract: Cloud computing is becoming increasingly popular due to its on-demand resource allocation and scalability. It is essential to precisely anticipate workload as applications and users on cloud-based services increase to distribute resources effectively and avoid service interruptions. We present an overview of approaches for workload forecasting in cloud systems in this study. We explore more sophisticated approaches like algorithms for deep learning (DL) and machine learning (ML) in addition to more conventional approaches like analysis of time series and models of regression. We also discuss difficulties and unresolved research questions in the area of workload forecasting for cloud settings. Cloud service providers may allocate resources wisely and guarantee good performance and accessibility for their clients by being aware of these techniques and problems. Cloud computing with virtualization and customized service is crucial to improving the service provided to customers. Accurate forecasting of workload is key to optimizing cloud performance. In this study, we discuss some methods of predicting workload in cloud environments. This study presents an overview of workload prediction techniques in cloud systems, with a special emphasis on long short-term memory (LSTM) networks. We go through the fundamental ideas behind LSTM networks and how well they can detect long-term relationships in data from time series. We also examine the particular difficulties and factors involved in LSTM-based workload forecasting implementation in cloud systems. We also examine previous research and methods that have employed LSTM networks to forecast workload in cloud systems. We examine the benefits and drawbacks of different methods, focusing on their effectiveness, scalability, and interpretability.
Abstract: The newly developed method of inpainting the missing region in digital images that is based on the isotropic model was applied to a variety of color spaces, including XYZ, YCbCr, NTSC, and HSV color spaces. Tests were performed using masks that the researcher had created based on the original (clear) images; these masks range in both size and shape. In order to evaluate the quality of image inpainting, statistical quality metrics such as MSE (Mean Square Error), PSNR (Peak Signal to Noise Ratio), and SSIM (Structural Similarity Index) are utilized. The performance of this method was superior to the state-of-the-art PDEs (partial differential equations) inpainting methods.
Abstract: In this paper, specific forms of non-linear systems of integral equations have been solved using the Sumudu transform and the Adomian decomposition method. To illustrate the method, three examples of various forms in this class of functional equations have been developed. The method has produced an approximate-exact solution for several systems. The method's accuracy, efficiency, and simplicity are demonstrated by the results of applying it to various systems of these kinds of integral equations.
Abstract: Internet of Thing (IoT)" is a major and emerging technology where different devices are associated together to operate smartly deprived of human’s intervention. IoT has a great effect on economic, social and commercial life. The extremely large number of associated devices together through different protocols make it vulnerable to be threatened by several types of attacks, such as; Unauthorized Access and Authentication Attacks, Network Attacks, Data Privacy and Integrity Attacks, and Supply Chain Attacks. Building and developing lightweight authentication systems for IoT devices are indeed crucial due to several compelling reasons, including; Resource Constraints, Security Concerns, Scalability, Bandwidth Efficiency, Latency Considerations, Cost Reduction, Long Lifecycle and the overall User Experience. Blockchain and “Machine Learning (ML)” are emerging technologies that may be exploited in solving different security problems. This paper presents a survey of lightweight authentication protocols and schemes that are adopted and applied for IoT environment. The security issues in IoT environment are also discussed besides the different types of attacks that may face the network. The paper held a comparison between the selected works and studies in terms of different criteria including the benefits, the results, the used applications and technologies or methods.
Abstract: The main aim of this work is to find new codes arising from construct a complete (k , n) – arcs in PG( 3,17), when n = 3,4,5 we take the union of some (k,n) – arcs. Furhtermore, when n = 6,7,8,…,307 by using matlab19B program (1) to found all construct a complete ( k , n + 1) - arcs from complete (k , n) - arcs. We start with points index zero and unit point, which we call the basic points of table (points and planes), then we start adding points from the remaining points of install the points of the first arc until we get intersections through which deleted all points of the plane , then we repeat that method until we get the maximum complete (5220 ; 307) - arcs. Then we find the [k, n, d] q-code of each a complete (k , n) - arcs .
Abstract: In this paper, the time-dependent Emden-Fowler type partial differential equations and wave-type equations with singular behavior at are analytically solved using the combined Laplace transform and Adomain decomposition method (LT-ADM). To avoid the singularity behavior for both models at , the benefit of this single global technique is used to present a solid framework. The method is shown to produce approximate-exact solutions to various kinds of problems in One-dimensional space. The results gained in each case demonstrate the dependability and effectiveness of this approach. To show the high accuracy of the approximate solution results (LT-ADM), compare the absolute errors obtained by the Padé approximation (PA) of order compared with the exact solution.
Abstract: Bifurcation theory is a field of mathematics that studies the qualitative changes in the behavior of a dynamical system as a parameter in the system is varied.In this work, we review the history, the types of bifurcations and the relations of those concepts with the chaos phenomena and some other concepts such as sensitivity to the initial value. We also survey some applications of this important theory in the other fieldsLike Physics, chemistry, biology, geography, … etc>
Abstract: This study presents a new algorithm for effectively solving the nonlinear fractional Korteweg-de Vries-Burger equation (NFKDV-B) using a hybrid explicit finite difference technique with the Adomian polynomial (HEFD). The suggested technique addresses the problem of accurately solving the FKDV-B equation with fractional nonlinear space derivatives in numerical solutions. Numerical results are obtained by comparing the exact solution with absolute and mean square errors. The fractional time and space derivatives are estimated using two widely used techniques: the Caputo derivative and the shifted Grünwald-Letnikov (G-L) formulas. Using a test problem to asses the HEFD method accuracy against the exact solution and the conventional explicit finite difference (EFD) method. The results exhibit excellent agreement between the approximate and exact solutions at different time values. The findings highlight the effectiveness of the proposed method across a range of fractional derivative values when compared to the exact solution and conventional explicit finite difference methods.
Abstract: Traditional (pixel-by-pixel) classification techniques are time-consuming, whereas semantic segmentation in machine learning requires assigning class labels to each pixel in an image. This study proposes a block-by-block (5x5 chunks) segmentation method for semantic segmentation, which involves image dissection, feature extraction, and model training based on specific color and textural properties. Thirty cat photos from the Oxford-IIIT Pet dataset were used for evaluation. Five different Artificial Neural Network (ANN) models, including LM, BGFGS, RP, SCG, and GDX, were trained and assessed for both pixel-based and block-based methods. The accuracy of the block-based classification ranges from 82.94% to 85.83%, surpassing the pixel-based approach, which ranges from 70.82% to 76.47%. The processing time for the models also improved with the block-based method. For the pixel-based approach, RP model takes the longest processing time i.e., 242.39 seconds, while GDX model takes the shortest processing time i.e., 49.89 seconds. For the block-based approach, LM model takes the longest processing time i.e., 13.86 seconds, while GDX still has the shortest processing time i.e., 5.98 seconds. Therefore, block-based methods can be seen as more efficient and accurate for classification models. The LM model achieved the highest accuracy on test images, ranging from 94.72% to 89.81%, while the GDX model had the lowest accuracy, ranging from 92.96% to 81.15%. The remaining models, RP, SCG, and BFG, have intermediate levels of accuracy.
Abstract: This paper introduces three members of a one-step optimized third derivative hybrid block method family for solving general second-order initial value problems. The methodology incorporates optimization into the derivation process to achieve enhanced accuracy. Through rigorous analysis, it is demonstrated that all the derived methods are found to be zero-stable, consistent, A-stable, and convergent. The implementation of these newly derived methods is validated through numerical experimentation, where the results exhibit superior accuracy compared to certain existing numerical methods explored in the study. All the newly derived optimized third derivative hybrid block methods possessed very small error constants with high order accuracy.
Abstract: The new distance defined on a connected graph G contains of three terms: The ordinary distance between any two vertices in G, both the sum and the product of the two vertices' degrees, as this distance is more useful than the ordinary distance, especially in chemical structures because of its effect on the number of bonds (edges) on the atoms (vertices) carbon (graph). In this article, were found index with respect to new distance (d- index) of regular graph, in addition, finding the relationships between d-index. Also, The relationships and The graph were found between the diameter and the radius for the ordinary distance and between the diameter and the radius for the new distance, and finally, The d-index was found for the join operation of two and nth graphs.
Abstract: Mathieu proved that any prime algebra which is also C* - algebra is an ultraprime. Brešar shows that C* - algebra is an ultrasemiprime. This paper is give condtions to ultrasempirime algebras to get ultraprime. Mathieu defined ultraprime algebras by using four equivalent conditions. In his definition of ultraprime algebra, he used the ultrapower of algebra. In contrast, in other identical conditions, he once used sequences; in another occasion, he used a particular linear operator. Our proof adds a new condition by using the prime algebra to the ultrasemiprime algebra to get an equivalent condition to ultraprime algebras.