Physics, Mathematics, Computer science

Thumbnail Image
Director
Valerie Le Guyon
Parent Organisation
Sorbonne University Abu Dhabi Library
Country
AE
City
Abu Dhabi

Filters

Reset filters

Settings

Now showing 1 - 10 of 173
  • Publication
    Optimum sensors allocation for drones multi-target tracking under complex environment using improved prairie dog optimization
    (2024)
    Abu Zitar, Raed 
    ;
    Alhadhrami, Esra Ebrahim 
    ;
    Abualigah, Laith
    ;
    Barbaresco, Frederic
    ;
    Seghrouchni, Amal ElFallah
    This paper presents a novel hybrid optimization method to solve the resource allocation problem for multi-target multi-sensor tracking of drones. This hybrid approach, the Improved Prairie Dog Optimization Algorithm (IPDOA) with the Genetic Algorithm (GA), utilizes the strengths of both algorithms to improve the overall optimization performance. The goal is to select a set of sensors based on norms of weighted distances cost function. The norms are the Euclidean distance and the Mahalanobis distance between the drone location and the sensors. The second one depends on the predicted covariance of the tracker. The Extended Kalman Filter (EKF) is used for state estimation with proper clutter and detection models. Since we use Multi-objects to track, the Joint Probability Distribution Function (JPDA) estimates the best measurement values with a preset gating threshold. The goal is to find a sensor or minimum set of sensors that would be enough to generate high-quality tracking based on optimum resource allocation. In the experimentation simulated with Stone Soup, one radar among five radars is selected at every time step of 50-time steps for 200 tracks distributed over 20 different ground truths. The proposed IPDOA provided optimum solutions for this complex problem. The obtained solution is an optimum offline solution that is used to select one or more sensors for any future flights within the vicinity of the 5 radars. Environment and conditions are assumed to be similar in future drone flights within the radars’ defined zone. The IPDOA performance was compared with the other 8 metaheuristic optimization algorithms and the testing showed its superiority over those techniques for solving this complex problem. The proposed simulated model can find the most relevant sensor(s) capable of generating the best quality tracks based on weighted distance criteria (Euclidean and Mahalanobis ). That would cut down the cost of operating extra sensors and then it would be possible to move them to other vicinity.
  • Publication
    Knowledge Fusion by Harnessing Support Vector Machines for Collaborative Uncertain Data Classification in Multiagent Systems
    (2024)
    Hussein, Ahmad MohdAziz
    ;
    Al-azzeh, Rashed M H
    ;
    Mughaid, Ala
    ;
    Abu Zitar, Raed 
    ;
    Migdady, Hazem
    ;
    Abualigah, Laith
    Distributed data mining (DDM) has emerged as a useful method for analyzing data that is spread across multiple sources. Nevertheless, DDM has other challenges that restrict its effectiveness, such as autonomy, privacy, efficiency, and implementation. DDM's rigidity and lack of adaptability may render it unsuitable for numerous applications due to its requirement for a consistent environment, administration, control, and categorization procedures. In order to address these challenges, we suggest the implementation of MAS-DDM, which combines a multiagent system (MAS) with DDM. MAS, or Multiagent Systems, is a methodology used to create independent agents that possess shared environments and can collaborate and communicate with one another. The study showcases the advantages and attractiveness of MAS-DDM. In the context of MAS-DDM, agents can exchange their thoughts, even when the data they possess is classified and cannot be disclosed. Other agents can then decide whether to incorporate these beliefs into their decision-making process, which may result in a revision of their initial assumptions about each data class. MAS-DDM focuses on the support vector machine (SVM) method, which is commonly employed for handling uncertain data. Our investigation demonstrates that the performance of MAS-DDM surpasses that of DDM strategies that do not incorporate communicative processes, even when all MAS-DDM agents utilize the same methodology. We present empirical evidence demonstrating that the precision of the categorization job is significantly enhanced through the exchange of knowledge among agents.
      3  1
  • Publication
    An investigation to identify the factors that cause failure in English essay, precis, and composition papers in CSS exams
    (2024)
    Gul, Kiran
    ;
    Shahzad, Waheed
    ;
    Raza, Ali
    ;
    Hanandeh, Essam
    ;
    Abu Zitar, Raed 
    ;
    Aldiabat, Khaled
    ;
    Shboul, Rabah
    ;
    Abualigah, Laith
    The research study aims to examine why candidates in Pakistan failed the English Essay, Precis, and Composition sections of the Central Superior Services (CSS) tests. Those candidates chosen for various civil service positions take the prestigious and difficult CSS exam. The study aims to discover candidates’ difficulties in these particular CSS exam sections and investigate methods for enhancing their English language ability. A mixed-methods strategy is used in the research process to collect both quantitative and qualitative data. Participants in the CSS exam who once took the English Essay, Precis, and Composition papers and got fail in it received a survey form to respond according to their experience. Other than this, we also conducted semi-structured interviews with CSS test winners currently working as officials, such as Deputy Commissioners, Assistant Commissioners, Assistant Superintendents of Police, and Deputy Superintendents of Police. Insights into the causes of failure and the experiences of successful candidates are sought after from both data sources. The research findings highlighted several key factors contributing to failure in English Essays, Precis, and Composition papers. These factors include lack of comprehension and understanding, grammatical errors, inadequate organization, poor handwriting, insufficient practice, lack of originality, difficulty in adapting to essay prompts and precis passages, poor organization, failure to understand and address the purpose, insufficient development of ideas, failure to reach the required word count, grammatical mistakes, neglecting proofreading and revision, poor writing expression, and weak induction and conclusion in essays, tough paper pattern old formatted curriculum. Participants reported struggling to express their ideas coherently, having limited language skills, facing challenges in managing time effectively, lacking proper precis structure understanding, inadequate expertise in the subject, lack of training and resources, lack of analytical and critical thinking abilities, inadequate exam preparation, time management issues, poor grammar abilities, exam phobia, and limited vocabulary as potential factors contributing to failure.
  • Publication
    Optimizing Aircraft Pitch Control Systems: A Novel Approach Integrating Artificial Rabbits Optimizer with PID-F Controller
    (2024)
    Abualigah, Laith
    ;
    Izci, Davut
    ;
    Ekinci, Serdar
    ;
    Abu Zitar, Raed 
    The precise control of aircraft pitch angles is critical in aviation for maintaining specific attitudes during flight, including straight and level flight, ascents, and descents. Traditional control strategies face challenges due to the non-linear and uncertain dynamics of flight. To address these issues, this study introduces a novel approach employing the artificial rabbits optimizer (ARO) for tuning a PID controller with a filtering mechanism (PID-F) in aircraft pitch control systems. This combination aims to enhance the stability and performance of the aircraft pitch control system by effectively mitigating the kick effect through the incorporation of a filter coefficient in the derivative gain. The study employs a time-domain-based objective function to guide the optimization process. Simulation results validate the stability and consistency of the proposed ARO/PID-F approach. Comparative analysis with various optimization algorithm-based controllers from the literature demonstrates the effectiveness of the proposed technique. Specifically, the ARO/PID-F controller exhibits a rapid response, zero overshoot, minimal settling time, and precise control during critical phases. The obtained results position the proposed methodology as a promising and innovative solution for optimizing aircraft pitch control systems, offering improved performance and reliability.
      5
  • Publication
    Modified Aquila Optimizer Feature Selection Approach and Support Vector Machine Classifier for Intrusion Detection System
    (2024)
    Abu Zitar, Raed 
    ;
    Abualigah, Laith
    ;
    Ahmed, Saba Hussein
    ;
    Almomani, Mohammad H.
    ;
    Alsoud, Anas Ratib
    ;
    Abuhaija, Belal
    ;
    Hanandeh, Essam Said
    ;
    Jia, Heming
    ;
    Abd Elminaam, Diaa Salama
    ;
    Abd Elaziz, Mohamed
    With the ever-expanding ubiquity of the Internet, wireless networks have permeated every facet of modern life, escalating concerns surrounding network security for users. Consequently, the demand for a robust Intrusion Detection System (IDS) has surged. The IDS serves as a critical bastion within the security framework, a significance further magnified in wireless networks where intrusions may stem from the deluge of sensor data. This influx of data, however, inevitably taxes the efficiency and computational speed of IDS. To address these limitations, numerous strategies for enhancing IDS performance have been posited by researchers. This paper introduces a novel feature selection method grounded in Support Vector Machine (SVM) and harnessing the innovative modified Aquila Optimizer (mAO) for Intrusion Detection Systems in Wireless Sensor Networks. To evaluate the efficacy of our approach, we employed the KDD'99 dataset for testing and benchmarking against established methods. Multiple performance metrics, including accuracy, detection rate, false alarm rate, feature count, and execution time, were utilized for assessment. Our comparative analysis reveals the superiority of the proposed method, with standout results in terms of feature reduction, detection accuracy, and false alarm mitigation, yielding significant improvements of 11%, 98.76%, and 0.02%, respectively.
      5
  • Publication
    Ten years of generative adversarial nets (GANs): a survey of the state-of-the-art
    (2024)
    Chakraborty, Tanujit 
    ;
    K S, Ujjwal Reddy
    ;
    Naik, Shraddha M
    ;
    Panja, Madhurima
    ;
    Manvitha, Bayapureddy
    Generative adversarial networks (GANs) have rapidly emerged as powerful tools for generating realistic and diverse data across various domains, including computer vision and other applied areas, since their inception in 2014. Consisting of a discriminative network and a generative network engaged in a minimax game, GANs have revolutionized the field of generative modeling. In February 2018, GAN secured the leading spot on the ‘Top Ten Global Breakthrough Technologies List’ issued by the Massachusetts Science and Technology Review. Over the years, numerous advancements have been proposed, leading to a rich array of GAN variants, such as conditional GAN, Wasserstein GAN, cycle-consistent GAN, and StyleGAN, among many others. This survey aims to provide a general overview of GANs, summarizing the latent architecture, validation metrics, and application areas of the most widely recognized variants. We also delve into recent theoretical developments, exploring the profound connection between the adversarial principle underlying GAN and Jensen–Shannon divergence while discussing the optimality characteristics of the GAN framework. The efficiency of GAN variants and their model architectures will be evaluated along with training obstacles as well as training solutions. In addition, a detailed discussion will be provided, examining the integration of GANs with newly developed deep learning frameworks such as transformers, physics-informed neural networks, large language models, and diffusion models. Finally, we reveal several issues as well as future research outlines in this field.
      4Scopus© Citations 1
  • Publication
    Drone Detection with Improved Precision in Traditional Machine Learning and Less Complexity in Single Shot Detectors
    (2024)
    Kassab, Mohamad
    ;
    Abu Zitar, Raed 
    ;
    Barbaresco, Frederic
    ;
    Seghrouchni, Amal El Fallah
    This work presents a broad study of drone detection based on a variety of machine-learning methods including traditional and deep-learning techniques. The data sets used are images obtained from sequences of video frames in both RGB and IR formats, filtered and unfiltered. First, traditional machine learning techniques such as SVM and RF were investigated to discover their drawbacks and study their feasibility in drone detection. It was evident that those techniques are not suitable for complex data sets (sets with several non-drone objects and clutter in the background). It was observed that the sliding window size results in a bias toward the selection of the bounding box when using the traditional NMS method. Therefore, to address this issue, a modified NMS is proposed and tested on SVM and RF. SVM and RF with modified NMS managed to achieve a relative improvement of up to 25% based on the evaluation metric. The Deep Learning techniques, on the other hand, showed better detection performance but less improvement when using the proposed NMS method. Since their biggest drawback is complexity, a modified deep learning paradigm was proposed to mitigate the usual complexity associated with deep learning methods. The proposed paradigm uses (Single Shot Detector) SSD and AdderNet filters in an attempt to avoid excessive multiplications in the convolutional layers. To demonstrate our method, the most common deep-learning techniques were comparatively tested to create a baseline for evaluating the proposed SSD/AdderNet. The training and testing of the deep learning models were repeated six times to investigate the consistency of learning in terms of parameters and performance. The proposed model was able to achieve better results with respect to the IR data set compared to its counterpart while reducing the number of multiplications at the convolutional layers by 43.42%. Moreover, and as a result of lower complexity, the proposed SSD/AdderNet showed fewer training and inference times compared to its counterpart.
      7
  • Publication
    Adapted arithmetic optimization algorithm for multi-level thresholding image segmentation: a case study of chest x-ray images
    (2023)
    Abu Zitar, Raed 
    ;
    Otair, Mohammad
    ;
    Abualigah, Laith
    ;
    Tawfiq, Saif
    ;
    Alshinwan, Mohammad
    ;
    Ezugwu, Absalom E.
    ;
    Sumari, Putra
    Particularly in recent years, there has been increased interest in determining the ideal thresholding for picture segmentation. The best thresholding values are found using various techniques, including Otsu and Kapur-based techniques. These techniques work well for bi-level thresholding, but when used to find the appropriate thresholds for multi-level thresholding, there will be issues with long calculation times, high computational costs, and the need for accuracy improvements. This work investigates the capability of the Arithmetic Optimization Algorithm to discover the best multilayer thresholding for picture segmentation to circumvent this issue. The leading mathematical arithmetic operators' distributional nature is used by the AOA method. The picture histogram was used to construct the candidate solutions in the modified algorithms, which were then updated according to the algorithm's features. The solutions are evaluated using Otsu's fitness function throughout the optimization process. The picture histogram is used to display the algorithm's potential solutions. The proposed approach is tested on five frequent photos from the Berkeley University database. The fitness function, root-mean-squared error, peak signal-to-noise ratio, and other widely used assessment metrics were utilized to assess the performance of the suggested segmentation approach. Many benchmark pictures were employed to verify the suggested technique's effectiveness and evaluate it against other well-known optimization methods described in the literature.
      16
  • Publication
    Autokeras Approach: A Robust Automated Deep Learning Network for Diagnosis Disease Cases in Medical Images
    (2023)
    Alaiad, Ahmed
    ;
    Migdady, Aya
    ;
    Al-Khatib, Raed M
    ;
    Alzoubi, Omar
    ;
    Abu Zitar, Raed 
    ;
    Abualigah, Laith
    Automated deep learning is promising in artificial intelligence (AI). However, a few applications of automated deep learning networks have been made in the clinical medical fields. Therefore, we studied the application of an open-source automated deep learning framework, Autokeras, for detecting smear blood images infected with malaria parasites. Autokeras is able to identify the optimal neural network to perform the classification task. Hence, the robustness of the adopted model is due to it not needing any prior knowledge from deep learning. In contrast, the traditional deep neural network methods still require more construction to identify the best convolutional neural network (CNN). The dataset used in this study consisted of 27,558 blood smear images. A comparative process proved the superiority of our proposed approach over other traditional neural networks. The evaluation results of our proposed model achieved high efficiency with impressive accuracy, reaching 95.6% when compared with previous competitive models.
      16Scopus© Citations 3
  • Publication
    Revealing Size‐Dependency of Ionic Liquid to Assist Perovskite Film Formation Mechanism for Efficient and Durable Perovskite Solar Cells
    (2023)
    Li, Liang 
    ;
    Wang, Fei
    ;
    Zhou, Kang
    ;
    Liang, Xiao
    ;
    Zhou, Xianfang
    ;
    Duan, Dawei
    ;
    Ge, Chuangye
    ;
    Zhang, Xintao
    ;
    Shi, Yumeng
    ;
    Lin, Haoran
    ;
    Zhu, Quanyao
    ;
    Hu, Hanlin
    ;
    Zhang, Hongyu
    Ionic liquids (ILs) are extensively utilized for the manipulation of crystallization kinetics of perovskite, morphology optimization, and defect passivation for the fabrication of highly efficient and stable devices. However, comparing ILs with different chemical structures and selecting the appropriate ILs from the many types available to enhance perovskite device performance remains a challenge. In this study, a range of ILs containing different sizes of anions are introduced as additives for assisting in film formation in perovskite photovoltaics. Specifically, ILs with various sizes significantly affects the strength of chemical interaction between ILs and perovskite composition, inducing varying degrees of conversion of lead iodide to perovskite as well as the formation of perovskite films with markedly disparate grain sizes and morphology. Theoretical calculations in conjunction with experimental measurements revealed that small-sized anion can more effectively reduce defect density by filling halide vacancies within perovskite bulk materials, resulting in suppression of charge-carrier recombination, an extended photoluminescence lifetime, and significantly improved device performance. Boosted by ILs with appropriate size, the champion power conversion efficiency of 24.09% for the ILs-treated device is obtained, and the unencapsulated devices retain 89.3% of its original efficiency under ambient conditions for 2000 h.
    Scopus© Citations 5  4