Article
Computer Science, Information Systems
Israa Al-Badarneh, Maria Habib, Ibrahim Aljarah, Hossam Faris
Summary: This paper introduces three stochastic and metaheuristic algorithms to train MLP neural network for solving the problem of imbalanced classifications. The algorithms are evaluated using accuracy, F-score, and G-mean, and the results show that F-score and G-mean are more advantageous when the datasets are imbalanced.
JOURNAL OF KING SAUD UNIVERSITY-COMPUTER AND INFORMATION SCIENCES
(2022)
Article
Computer Science, Artificial Intelligence
Divya Bairathi, Dinesh Gopalani
Summary: The proposed improved salp swarm algorithm, by integrating multiple elements, enhances exploration and exploitation capabilities, making it more effective in solving complex multimodal problems.
Article
Computer Science, Information Systems
Archana R. Panhalkar, Dharmpal D. Doye
Summary: The study proposes the creation of globally optimized decision trees using the African Buffalo Optimization algorithm and validates it through experiments, showing an improvement in accuracy and reduced decision tree size.
JOURNAL OF KING SAUD UNIVERSITY-COMPUTER AND INFORMATION SCIENCES
(2022)
Article
Mathematics
Nebojsa Bacanin, Ruxandra Stoean, Miodrag Zivkovic, Aleksandar Petrovic, Tarik A. Rashid, Timea Bezdan
Summary: An enhanced version of the firefly algorithm was proposed in this paper, addressing the drawbacks of the original method through an exploration mechanism and local search strategy. This algorithm was validated for selecting the optimal dropout rate for deep neural network regularization and also applied in image processing tasks.
Article
Computer Science, Artificial Intelligence
Mohammed Alweshah
Summary: This paper introduces the application of classification technique, coronavirus herd immunity optimizer algorithm, probabilistic neural network, and their effectiveness in solving classification problems. The experimental results show that the proposed CHIO-PNN method achieves high classification accuracy on multiple datasets, and it converges faster.
Article
Chemistry, Multidisciplinary
Kuan Wang, Qing Hu, Bin Gao, Qi Lin, Fu-Wei Zhuge, Da-You Zhang, Lun Wang, Yu-Hui He, Ralph H. Scheicher, Hao Tong, Xiang-Shui Miao
Summary: This study utilizes a novel CuS/GeSe conductive-bridge threshold switching memristor to fabricate electronic stochastic neurons, which enables Bayesian inference and improves accuracy in tumor diagnosis tasks compared to conventional artificial neural networks. Stochastic neurons, as opposed to deterministic ones, allow for better estimation of uncertainty and improve judgement accuracy by 81.2% in spiking neural networks.
MATERIALS HORIZONS
(2021)
Article
Automation & Control Systems
Ouail Mjahed, Salah El Hadaj, El Mahdi El Guarmah, Soukaina Mjahed
Summary: This paper investigates the optimization effects of multiple bio-inspired metaheuristic algorithms on hybrid ANNs, proposing a method that takes into account different datasets. Through classification experiments on four datasets, the results show that networks based on PSO-ANN have higher efficiency values.
STUDIES IN INFORMATICS AND CONTROL
(2022)
Article
Automation & Control Systems
Ouail Mjahed, Salah El Hadaj, El Mahdi El Guarmah, Soukaina Mjahed
Summary: This paper investigates the use of multiple hybrid artificial neural networks (ANNs) to classify different types of datasets, and optimizes the efficiency of the neural networks using multiple bio-inspired metaheuristic algorithms. The results show that the metaheuristic algorithms can achieve optimal efficiency, with PSO-ANN-based networks performing the best.
STUDIES IN INFORMATICS AND CONTROL
(2022)
Article
Multidisciplinary Sciences
Sultan Zeybek, Duc Truong Pham, Ebubekir Koc, Aydin Secer
Summary: A novel metaheuristic optimization approach for training deep RNNs for sentiment classification task is proposed, demonstrating significant advantages. The BA-3+ algorithm outperforms DE and PSO algorithms in terms of speed and accuracy, surpassing SGD, DE, and PSO algorithms.
Article
Computer Science, Artificial Intelligence
Francesco Ponzio, Enrico Macii, Elisa Ficarra, Santa Di Cataldo
Summary: In real-world scenarios, training Convolutional Neural Networks (CNNs) with high quality images and correct labels is difficult. This affects the performance of CNNs during both training and inference. To tackle this issue, we propose a new two-module CNN called Wise2WipedNet (W2WNet), which uses Bayesian inference to identify and discard spurious images during training and provides prediction confidence during inference. Our experiments on various image classification tasks and histological image analysis demonstrate that W2WNet can effectively identify image degradation and mislabelling issues, resulting in improved classification accuracy.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Computer Science, Artificial Intelligence
Edgar S. Garcia-Trevino, Pu Yang, Javier A. Barria
Summary: This article proposes a novel wavelet probabilistic neural network (WPNN) that relies on wavelet-based estimation of class probability densities for generative learning. The new neural network approach overcomes the limitation of traditional probabilistic neural networks by not being dependent on the number of data inputs. It is particularly suitable for data stream classification and anomaly detection in offline and online environments, and achieves significant performance improvements compared to existing algorithms.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Abdesslem Layeb
Summary: This article introduces a new population-based optimization algorithm called Tangent Search Algorithm (TSA) for solving optimization problems. The TSA utilizes a mathematical model based on the tangent function to move a given solution towards a better solution, balancing between exploitation and exploration search. It also incorporates a novel escape procedure to avoid local minima and an adaptive variable step-size for enhanced convergence capacity. Experimental results show that the TSA algorithm yields promising and competitive results in various tests, demonstrating its simplicity, efficiency, and requirement of only a small number of user-defined parameters.
NEURAL COMPUTING & APPLICATIONS
(2022)
Article
Computer Science, Theory & Methods
El-Ghazali Talbi
Summary: This article proposes a unified way to describe various optimization algorithms, focusing on common and important search components. This methodology has also been extended to advanced optimization approaches including surrogate-based, multi-objective, and parallel optimization.
ACM COMPUTING SURVEYS
(2021)
Article
Computer Science, Hardware & Architecture
B. Jaishankar, Raghunathan Anitha, Finney Daniel Shadrach, M. Sivarathinabala, V Balamurugan
Summary: Categorizing music files according to their genre is a challenging task in Music Information Retrieval (MIR). This study proposes a method for music genre classification using features extracted from audio data, with feature selection performed using African Buffalo Optimization (ABO) and classification carried out using various classifiers. The results show that the ABO based feature selection strategy achieves an average accuracy of 82% with a mean square error (MSE) of 0.003 when used with a neural network classifier.
COMPUTER SYSTEMS SCIENCE AND ENGINEERING
(2023)
Article
Mathematics
Nuria Molla, Alejandro Rabasa, Jesus J. Rodriguez-Sala, Joaquin Sanchez-Soriano, Antonio Ferrandiz
Summary: Data science is a promising field that supports the decision-making process. Data streams provide updated knowledge for decision support systems. The Incremental Decision Rules Algorithm (IDRA) is a new approach that improves accuracy in decision support systems. Experimental results demonstrate that IDRA outperforms VFDR and CREA in most scenarios.
Article
Computer Science, Artificial Intelligence
Mohammed Alweshah, Saleh Al Khalaileh, Brij B. Gupta, Ammar Almomani, Abdelaziz Hammouri, Mohammed Azmi Al-Betar
Summary: This study combined the latest Monarch Butterfly Optimization algorithm with a wrapper feature selection method using k-nearest neighbor classifier, and conducted experiments on 18 benchmark datasets. The results showed that the MBO algorithm was superior to other algorithms in terms of high classification accuracy rate and reduced selection size.
NEURAL COMPUTING & APPLICATIONS
(2022)
Article
Computer Science, Artificial Intelligence
Malik Braik, Mohammad Hashem Ryalat, Hussein Al-Zoubi
Summary: This paper presents a novel meta-heuristic algorithm called Ali Baba and the forty thieves (AFT) for solving global optimization problems. The AFT algorithm was evaluated on basic benchmark test functions and challenging benchmarks, showing stronger performance than other well-studied algorithms. The applicability of the AFT algorithm was demonstrated in solving engineering design problems, further emphasizing its potential performance.
NEURAL COMPUTING & APPLICATIONS
(2022)
Article
Computer Science, Artificial Intelligence
Malik Braik, Hussein Al-Zoubi, Mohammad Ryalat, Alaa Sheta, Omar Alzubi
Summary: Crow Search Algorithm (CSA) is a promising meta-heuristic method that mimics the intelligent behavior of crows in nature. By combining it with Particle Swarm Optimization (PSO), the Memory based Hybrid CSA (MHCSA) achieves a stronger diversity ability and a better balance between exploration and exploitation, effectively overcoming the early convergence and imbalance issues. Test results have demonstrated the superiority of MHCSA over other methods in terms of accuracy and stability.
ARTIFICIAL INTELLIGENCE REVIEW
(2023)
Article
Computer Science, Artificial Intelligence
Ammar Almomani
Summary: Virtual Private Networks (VPNs) are commonly used encrypted communication services for bypassing censorship and accessing geographically locked services. This study analyzed VPN and non-VPN traffic and developed a classification system using machine learning classifiers. The results showed that the proposed method accurately differentiated between VPN and non-VPN traffic with an accuracy level of approximately 99%.
EGYPTIAN INFORMATICS JOURNAL
(2022)
Article
Computer Science, Artificial Intelligence
Ankit K. Jain, Brij B. Gupta, Kamaljeet Kaur, Piyush Bhutani, Wadee Alhalabi, Ammar Almomani
Summary: Smishing is a cyber-security threat in which malicious text messages or SMS are sent to mobile users, causing significant financial losses. This paper proposes an efficient approach that analyzes text content and URLs in SMS to detect smishing messages. By integrating multiple classifiers, the proposed approach achieves a high accuracy and precision rate. The experimental results show that the proposed approach outperforms existing models in terms of efficiency and accuracy.
INTERNATIONAL JOURNAL OF INTELLIGENT SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Mohammad Hashem Ryalat, Osama Dorgham, Sara Tedmori, Zainab Al-Rahamneh, Nijad Al-Najdawi, Seyedali Mirjalili
Summary: Digital image processing techniques and algorithms are used to support medical experts in disease identification, studies, and diagnosis. Image segmentation methods are widely used in this area to simplify image representation and analysis. Among various approaches, multilevel thresholding methods have shown better results. However, traditional statistical approaches like the Otsu and Kapur methods suffer from high computational costs for multilevel thresholding. In this work, the Harris hawks optimization technique is combined with Otsu's method to reduce computational costs while maintaining optimal outcomes.
NEURAL COMPUTING & APPLICATIONS
(2023)
Article
Computer Science, Artificial Intelligence
Kwok Tai Chui, Brij B. Gupta, Rutvij H. Jhaveri, Hao Ran Chi, Varsha Arya, Ammar Almomani, Ali Nauman
Summary: This paper proposes a multiround transfer learning and modified generative adversarial network (MTL-MGAN) algorithm for lung cancer detection. The algorithm maximizes transferability through a multiround transfer learning process and avoids negative transfer through customized loss functions. The proposed algorithm significantly improves accuracy compared to related works.
INTERNATIONAL JOURNAL OF INTELLIGENT SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Meena Malik, Chander Prabha, Punit Soni, Varsha Arya, Wadee Alhalabi Alhalabi, Brij B. Gupta, Aiiad A. Albeshri, Ammar Almomani
Summary: Machine learning and deep learning have diverse applications in areas such as education, genetics, and space engineering. The development of smart cities is still in its early stage, with a major challenge being effective waste management. Smart city experts play a crucial role in formulating efficient waste management schemes that can be integrated into the city's overall development plan. This study proposes an automated classification model for urban waste using Convolutional Neural Networks, which can be implemented with low-cost resources and scaled up to address the issue of healthy living provisions across cities. The key aspects of developing such models involve using pre-trained models and employing transfer learning for fine-tuning.
INTERNATIONAL JOURNAL ON SEMANTIC WEB AND INFORMATION SYSTEMS
(2023)
Article
Business
Ammar Almomani
Summary: This paper proposes a new method using machine learning to analyze and classify darknet traffic, which demonstrates high accuracy and precision in distinguishing between malicious and benign traffic.
INFORMATION SYSTEMS AND E-BUSINESS MANAGEMENT
(2023)
Review
Computer Science, Information Systems
Kwok Tai Chui, Brij B. Gupta, Jiaqi Liu, Varsha Arya, Nadia Nedjah, Ammar Almomani, Priyanka Chaurasia
Summary: The smart city vision has driven the rapid development of IoT and CPS. This paper surveys various aspects of IoT and CPS from 2013 to May 2023. It discusses industry standards, big data utilization, machine learning algorithms, advanced security techniques, and research challenges. The focus has shifted towards advanced algorithms, such as deep learning, transfer learning, and data generation algorithms, to provide more accurate models.
Article
Computer Science, Theory & Methods
Mohammad Hashem Ryalat, Hussam N. Fakhouri, Jamal Zraqou, Faten Hamad, Mamon S. Alzboun, Ahmad K. Al Hwaitat
Summary: Data testing is crucial in software development, and this study proposes an improved version of the Multi-verse Optimizer called TMVO. TMVO considers the movement of the swarm and the mean of the two best solutions in the universe, ensuring efficient exploration and exploitation. It is applied to automatically develop test cases for structural data testing, with a focus on automating the data collection process. Tested on various functions and challenging programs, TMVO outperformed the original MVO algorithm in most cases.
INTERNATIONAL JOURNAL OF ADVANCED COMPUTER SCIENCE AND APPLICATIONS
(2023)
Article
Computer Science, Information Systems
Ammar Almomani, Mohammed Alweshah, Waleed Alomoush, Mohammad Alauthman, Aseel Jabai, Anwar Abbass, Ghufran Hamad, Meral Abdalla, Brij B. Gupta
Summary: Voice classification is essential for creating intelligent systems that assist with student exams, criminal identification, and security systems. The research aims to develop a system that can predict and classify gender, age, and accent, resulting in the proposal of a new system called Classifying Voice Gender, Age, and Accent (CVGAA). By incorporating rhythm-based features and using backpropagation and bagging algorithms, the voice recognition system's accuracy is significantly improved, with the Bagging algorithm achieving the highest accuracy of 55.39% in the voice common dataset and 78.94% in speech accent for age classification and accent accuracy.
CMC-COMPUTERS MATERIALS & CONTINUA
(2023)
Article
Computer Science, Interdisciplinary Applications
Musa Alshawabkeh, Mohammad Hashem Ryalat, Osama M. Dorgham, Khalid Alkharabsheh, Mohammad Hjouj Btoush, Mamoun Alazab
Summary: In this paper, a deep-learning approach is proposed for detecting diabetic retinopathy. The proposed approach combines image augmentation, contrast limited adaptive histogram equalisation, CNN and transfer learning, and ensemble classification, achieving high accuracy and stability compared to other approaches.
INTERNATIONAL JOURNAL OF COMPUTER APPLICATIONS IN TECHNOLOGY
(2022)
