Malek Algabri
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Malek Algabri has a total of 20 co-authors. Here are the top co-authors by number of shared publications:
- Saida M. Alkurkushi
Sana'a University, Sana'a
3 shared publications - Ayman Alsabry
International University of Technology Twintech
2 shared publications Mossa Ghurab
Faculty of Computer and Information Technology (FCIT), Sanaa University, Sanaa, Yemen
2 shared publicationsAbdualmajed A. G. Al-Khulaidi
Department of Computer Science, Faculty of Computer and Information Technology, Sana'a University, Sana'a, Yemen
1 shared publication- Abdualmajed Ahmed Ghaleb Al-Khulaidi
Faculty of Computer and Information Technology (FCIT), Sanaa University, Sanaa, Yemen
1 shared publication - Abdulsalam Tonin
Faculty of Computer and Information Technology (FCIT), Sanaa University, Sanaa, Yemen
1 shared publication
Latest Publications
- Open Access
- 3 September 2026
- Published by Springer Nature in Scientific Reports
Hybrid encoder-decoder to preserve long-range context in brain tumor segmentation with EfficientNetB7 and a convolutional vision transformer mechanism
- 3 September 2026
- Published by Springer Nature in Scientific Reports
Open AccessHybrid encoder-decoder to preserve long-range context in brain tumor segmentation with EfficientNetB7 and a convolutional vision transformer mechanism
- Open Access
- 7 January 2026
- Published by Springer Nature in Research Square
MICT‑Discord: Thresholds, Scaling, Noise Resilience, and Geometry
Quantum correlations extend beyond entanglement and play a central role in quantum information. Quantum discord captures nonclassical correlations that can persist in separable states. We present the first systematic geometric and operational analysis of MICT-discord in two-qubit systems, integrating numerical validation, scaling laws, noise robustness, and resource geometry. We validate MICT-discord against analytic benchmarks, uncover distinct scaling for entangled versus separable classes, and identify a crossover regime where separable states exceed entangled ones in discord, challenging monotonic assumptions. Under depolarizing and amplitude-damping channels, MICT-discord shows channel-dependent resilience, highlighting potential utility for NISQ devices. In contrast, maximally entangled mixed states (MEMS) expose limitations where MICT-discord fails to detect correlations at low purity. Geometric resource analysis delineates the convex hull boundaries of separable states and reveals the clustering of entangled states in high-discordance regions, providing operational thresholds. Overall, MICT-discord emerges as a structured, coherence-sensitive measure with demonstrated strengths in scaling and noise scenarios, alongside clear limitations in mixed-state regimes. Our results establish a reproducible, geometry-informed framework for discord analysis and suggest practical pathways for resource classification and benchmarking in quantum.
- 7 January 2026
- Published by Springer Nature in Research Square
Open AccessMICT‑Discord: Thresholds, Scaling, Noise Resilience, and Geometry
Quantum correlations extend beyond entanglement and play a central role in quantum information. Quantum discord captures nonclassical correlations that can persist in separable states. We present the first systematic geometric and operational analysis of MICT-discord in two-qubit systems, integrating numerical validation, scaling laws, noise robustness, and resource geometry. We validate MICT-discord against analytic benchmarks, uncover distinct scaling for entangled versus separable classes, and identify a crossover regime where separable states exceed entangled ones in discord, challenging monotonic assumptions. Under depolarizing and amplitude-damping channels, MICT-discord shows channel-dependent resilience, highlighting potential utility for NISQ devices. In contrast, maximally entangled mixed states (MEMS) expose limitations where MICT-discord fails to detect correlations at low purity. Geometric resource analysis delineates the convex hull boundaries of separable states and reveals the clustering of entangled states in high-discordance regions, providing operational thresholds. Overall, MICT-discord emerges as a structured, coherence-sensitive measure with demonstrated strengths in scaling and noise scenarios, alongside clear limitations in mixed-state regimes. Our results establish a reproducible, geometry-informed framework for discord analysis and suggest practical pathways for resource classification and benchmarking in quantum.
- Open Access
- 27 December 2025
- Published by MDPI AG in Brain Sciences
- Vol. 16 (1) , 37
A Multi-Task EfficientNetV2S Approach with Hierarchical Hybrid Attention for MRI Enhancing Brain Tumor Segmentation and Classification
Background: Brain tumors present a significant clinical problem due to high mortality and strong heterogeneity in size, shape, location, and tissue characteristics, complicating reliable MRI analysis. Existing automated methods are limited by non-selective skip connections that propagate noise, axis-separable attention modules that poorly integrate channel and spatial cues, shallow encoders with insufficiently discriminative features, and isolated optimization of segmentation or classification tasks. Methods: We propose a model using an EfficientNetV2S backbone with a Hierarchical Hybrid Attention (HHA) mechanism. The HHA couples a global-context pathway with a local-spatial pathway, employing a correlation-driven, per-pixel fusion gate to explicitly model interactions between them. Multi-scale dilated blocks are incorporated to enlarge the effective receptive field. The model is applied to a multiclass brain tumor MRI dataset, leveraging shared representation learning for joint segmentation and classification. Results: The design attains a Dice score of 92.25% and a Jaccard index of 86% for segmentation. For classification, it achieves an accuracy of 99.53%, with precision, recall, and F1 scores all close to 99%. These results indicate sharper tumor boundaries, stronger noise suppression in segmentation, and more robust discrimination in classification. Conclusions: The proposed framework effectively overcomes key limitations in brain tumor MRI analysis. The integrated HHA mechanism and shared representation learning yield superior segmentation quality with enhanced boundary delineation and noise suppression, alongside highly accurate tumor classification, demonstrating strong clinical utility.
- 27 December 2025
- Published by MDPI AG in Brain Sciences
- Vol. 16 (1) , 37
Open AccessA Multi-Task EfficientNetV2S Approach with Hierarchical Hybrid Attention for MRI Enhancing Brain Tumor Segmentation and Classification
Background: Brain tumors present a significant clinical problem due to high mortality and strong heterogeneity in size, shape, location, and tissue characteristics, complicating reliable MRI analysis. Existing automated methods are limited by non-selective skip connections that propagate noise, axis-separable attention modules that poorly integrate channel and spatial cues, shallow encoders with insufficiently discriminative features, and isolated optimization of segmentation or classification tasks. Methods: We propose a model using an EfficientNetV2S backbone with a Hierarchical Hybrid Attention (HHA) mechanism. The HHA couples a global-context pathway with a local-spatial pathway, employing a correlation-driven, per-pixel fusion gate to explicitly model interactions between them. Multi-scale dilated blocks are incorporated to enlarge the effective receptive field. The model is applied to a multiclass brain tumor MRI dataset, leveraging shared representation learning for joint segmentation and classification. Results: The design attains a Dice score of 92.25% and a Jaccard index of 86% for segmentation. For classification, it achieves an accuracy of 99.53%, with precision, recall, and F1 scores all close to 99%. These results indicate sharper tumor boundaries, stronger noise suppression in segmentation, and more robust discrimination in classification. Conclusions: The proposed framework effectively overcomes key limitations in brain tumor MRI analysis. The integrated HHA mechanism and shared representation learning yield superior segmentation quality with enhanced boundary delineation and noise suppression, alongside highly accurate tumor classification, demonstrating strong clinical utility.
- Open Access
- 5 December 2025
- Published by Springer Nature in Research Square
Discord-Enhanced Quantum Key Distribution in Noisy Mixed-State Channels
Quantum key distribution (QKD) facilitates secure communication by harnessing non-classical correlations within quantum systems. Although entanglement is fundamental to many QKD protocols, it is highly susceptible to noise and decoherence effects. In this work, we analyze the performance of Werner and Bell-diagonal states within commonly used QKD protocols such as BB84, Six-state, and their device-independent variants, considering realistic noise environments and eavesdropping strategies. Employing a Qiskit-based simulation toolkit , we assess critical performance metrics including quantum bit error rate (QBER), fidelity, concurrence, quantum discord, violations of the CHSH inequality, and Devetak, Winter key rates. Our findings emphasize the resilience of quantum discord as a resource in scenarios where entanglement is diminished, and reveal that Bell-diagonal states may exhibit superior performance compared to Werner states under certain noise conditions. We further discuss practical consequences for secure key generation and propose directions for extending this work toward physical device implementation and machine learning–enhanced detection of eavesdropping attempts. These results shed light on the robustness of QKD protocols that utilize mixed quantum states and the significant role quantum discord can play in secure quantum communication.
- 5 December 2025
- Published by Springer Nature in Research Square
Open AccessDiscord-Enhanced Quantum Key Distribution in Noisy Mixed-State Channels
Quantum key distribution (QKD) facilitates secure communication by harnessing non-classical correlations within quantum systems. Although entanglement is fundamental to many QKD protocols, it is highly susceptible to noise and decoherence effects. In this work, we analyze the performance of Werner and Bell-diagonal states within commonly used QKD protocols such as BB84, Six-state, and their device-independent variants, considering realistic noise environments and eavesdropping strategies. Employing a Qiskit-based simulation toolkit , we assess critical performance metrics including quantum bit error rate (QBER), fidelity, concurrence, quantum discord, violations of the CHSH inequality, and Devetak, Winter key rates. Our findings emphasize the resilience of quantum discord as a resource in scenarios where entanglement is diminished, and reveal that Bell-diagonal states may exhibit superior performance compared to Werner states under certain noise conditions. We further discuss practical consequences for secure key generation and propose directions for extending this work toward physical device implementation and machine learning–enhanced detection of eavesdropping attempts. These results shed light on the robustness of QKD protocols that utilize mixed quantum states and the significant role quantum discord can play in secure quantum communication.
- Open Access
- 9 September 2025
- Published by Springer Nature in Research Square
Quantum Correlations and Discord in Werner States for Quantum Teleportation
Quantum teleportation fundamentally depends on shared quantum correlations between spatially separated parties. Although entanglement is widely recognized as the primary resource enabling teleportation, quantum discord represents a more general form of nonclassical correlations that can persist even when entanglement is absent. We provide a thorough analysis combining analytical and numerical evaluations of quantum discord, negativity, and concurrence across varying noise parameters. The study includes simulations under both ide-alized and realistic noisy conditions, incorporating depolarizing noise models alongside experimentally relevant constraints. To improve quantum fidelity and accurately characterize quantum correlations in noisy environments, advanced error mitigation techniques—such as measurement error correction and zero-noise extrapolation—are applied. In this study, we focus on teleportation protocols utilizing two-qubit Werner states, which constitute a significant class of noisy mixed states modeling imperfect entanglement. We systematically examine how the teleportation fidelity varies as a function of the Werner mixing parameter λ and investigate its relationship with the corresponding quantum discord. Our findings indicate that quantum discord remains strictly greater than zero even in parameter regions where entanglement vanishes, emphasizing the role of discord in sustaining quantum advantages in teleportation schemes exposed to noise and imperfections. This highlights the importance of considering non-entanglement quantum correlations for a comprehensive understanding of teleportation resources in realistic settings.
- 9 September 2025
- Published by Springer Nature in Research Square
Open AccessQuantum Correlations and Discord in Werner States for Quantum Teleportation
Quantum teleportation fundamentally depends on shared quantum correlations between spatially separated parties. Although entanglement is widely recognized as the primary resource enabling teleportation, quantum discord represents a more general form of nonclassical correlations that can persist even when entanglement is absent. We provide a thorough analysis combining analytical and numerical evaluations of quantum discord, negativity, and concurrence across varying noise parameters. The study includes simulations under both ide-alized and realistic noisy conditions, incorporating depolarizing noise models alongside experimentally relevant constraints. To improve quantum fidelity and accurately characterize quantum correlations in noisy environments, advanced error mitigation techniques—such as measurement error correction and zero-noise extrapolation—are applied. In this study, we focus on teleportation protocols utilizing two-qubit Werner states, which constitute a significant class of noisy mixed states modeling imperfect entanglement. We systematically examine how the teleportation fidelity varies as a function of the Werner mixing parameter λ and investigate its relationship with the corresponding quantum discord. Our findings indicate that quantum discord remains strictly greater than zero even in parameter regions where entanglement vanishes, emphasizing the role of discord in sustaining quantum advantages in teleportation schemes exposed to noise and imperfections. This highlights the importance of considering non-entanglement quantum correlations for a comprehensive understanding of teleportation resources in realistic settings.
- Open Access
- 7 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 117771-117787
A Novel Energy Model for Optimizing Energy Efficiency and Prolonging Network Lifetime in MANET
Mobile ad-hoc networks are functional for usage, which require inter-node contact, including disaster relief, protection, and combat spotting. These networks face critical challenges, including rapid energy consumption in battery-limited devices and random topological changes resulting from node mobility, which lead to communication instability and packet loss. These factors reduce network efficiency and shorten lifespan, particularly in mission-critical applications such as environmental monitoring or military operations. This research addresses the imbalance in energy consumption caused by a node failure that acts as a relay node. It presents the Replacing Relay Nodes to Prolong Lifetime and improve Energy Efficiency (RePLEE) model. Comprehensive simulations conducted in NS-2 reveal that RePLEE significantly improves performance compared to standard AODV and enhanced EEAODV protocols. Specifically, RePLEE achieves a 13% reduction in energy consumption, a 61% decrease in packet loss, and a 26% reduction in end-to-end delay relative to AODV. Additionally, it extends network lifetime by 49% over AODV and by 26% over EEAODV. RePLEE also demonstrates superior scalability by maintaining high packet delivery ratios and lower control overhead across varying node densities. These results confirm the effectiveness of RePLEE in addressing the limitations of traditional protocols by enhancing energy efficiency, prolonging operational lifespan, and ensuring stable connectivity in MANET environments.
- 7 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 117771-117787
Open AccessA Novel Energy Model for Optimizing Energy Efficiency and Prolonging Network Lifetime in MANET
Mobile ad-hoc networks are functional for usage, which require inter-node contact, including disaster relief, protection, and combat spotting. These networks face critical challenges, including rapid energy consumption in battery-limited devices and random topological changes resulting from node mobility, which lead to communication instability and packet loss. These factors reduce network efficiency and shorten lifespan, particularly in mission-critical applications such as environmental monitoring or military operations. This research addresses the imbalance in energy consumption caused by a node failure that acts as a relay node. It presents the Replacing Relay Nodes to Prolong Lifetime and improve Energy Efficiency (RePLEE) model. Comprehensive simulations conducted in NS-2 reveal that RePLEE significantly improves performance compared to standard AODV and enhanced EEAODV protocols. Specifically, RePLEE achieves a 13% reduction in energy consumption, a 61% decrease in packet loss, and a 26% reduction in end-to-end delay relative to AODV. Additionally, it extends network lifetime by 49% over AODV and by 26% over EEAODV. RePLEE also demonstrates superior scalability by maintaining high packet delivery ratios and lower control overhead across varying node densities. These results confirm the effectiveness of RePLEE in addressing the limitations of traditional protocols by enhancing energy efficiency, prolonging operational lifespan, and ensuring stable connectivity in MANET environments.
- Open Access
- 5 March 2025
- Published by Thamar University in Journal of Engineering and Technological Sciences - JOEATS
- Vol. 3 (1) , 91-100
Approach to Plagiarism Detection in Programming Assignments
People tend to shortcut ways that save them time and effort to do the tasks required by them, either by taking tasks ready-made online, or stealing someone’s work as their own. Since everything now is connected to the Internet, there is a very high potential of duplicating or stealing someone else’s work, which is known as plagiarism. With the advancement of technology, it has become quite simple to do all tasks through the Internet. Plagiarism is the copying of other people’s ideas and actions; it is considered a crime. Plagiarism occurs due to laziness, fear of failure, and the desire to perform the required tasks without fatigue or effort. In this paper, a methodology for detecting plagiarism in programming tasks, in particular in the visual programming category using deep learning and machine learning algorithms is proposed. Also, a solution has been proposed to detect plagiarism in the source code and interfaces that pertain to programming assignments.
- 5 March 2025
- Published by Thamar University in Journal of Engineering and Technological Sciences - JOEATS
- Vol. 3 (1) , 91-100
Open AccessApproach to Plagiarism Detection in Programming Assignments
People tend to shortcut ways that save them time and effort to do the tasks required by them, either by taking tasks ready-made online, or stealing someone’s work as their own. Since everything now is connected to the Internet, there is a very high potential of duplicating or stealing someone else’s work, which is known as plagiarism. With the advancement of technology, it has become quite simple to do all tasks through the Internet. Plagiarism is the copying of other people’s ideas and actions; it is considered a crime. Plagiarism occurs due to laziness, fear of failure, and the desire to perform the required tasks without fatigue or effort. In this paper, a methodology for detecting plagiarism in programming tasks, in particular in the visual programming category using deep learning and machine learning algorithms is proposed. Also, a solution has been proposed to detect plagiarism in the source code and interfaces that pertain to programming assignments.
- Open Access
- 31 January 2024
- Published by Sana'a University in مجلة جامعة صنعاء للعلوم التطبيقية والتكنولوجيا
- Vol. 2 (1) , 7-15
Fake News Detection On Social Media:Review of Literature
In today's world, social media has become one of the most accessible sources of news for people worldwide due to its low cost, easy accessibility, and rapid dissemination. However, the harmful effects of fake news on individuals and society have been evident for a long time. Despite the challenges in detecting false information, numerous scholars are working to understand the issue and its characteristics. The purpose of this paper is to analyze contemporary frameworks that make use of different machine learning techniques in order to further our understanding of false news identification. The efficiency of these frameworks in recognizing and halting the spread of false information may be compared by examining various methods. A particular technique was used in the review process to provide a thorough and instructive assessment. Even though a lot of research has been done to identify false news, more can be done in the future. These studies have many limitations such as biased data, adaptability, credibility analysis and the news is only categorized as legitimate or unauthentic by their solutions. Nevertheless, a system of ratings or scores to assess the reliability of news is necessary for a workable approach.
- 31 January 2024
- Published by Sana'a University in مجلة جامعة صنعاء للعلوم التطبيقية والتكنولوجيا
- Vol. 2 (1) , 7-15
Open AccessFake News Detection On Social Media:Review of Literature
In today's world, social media has become one of the most accessible sources of news for people worldwide due to its low cost, easy accessibility, and rapid dissemination. However, the harmful effects of fake news on individuals and society have been evident for a long time. Despite the challenges in detecting false information, numerous scholars are working to understand the issue and its characteristics. The purpose of this paper is to analyze contemporary frameworks that make use of different machine learning techniques in order to further our understanding of false news identification. The efficiency of these frameworks in recognizing and halting the spread of false information may be compared by examining various methods. A particular technique was used in the review process to provide a thorough and instructive assessment. Even though a lot of research has been done to identify false news, more can be done in the future. These studies have many limitations such as biased data, adaptability, credibility analysis and the news is only categorized as legitimate or unauthentic by their solutions. Nevertheless, a system of ratings or scores to assess the reliability of news is necessary for a workable approach.
- Open Access
- 11 January 2024
- Published by SPIIRAS in Informatics and Automation
- Vol. 23 (1) , 129-168
Iterative Tuning of Tree-Ensemble-Based Models' parameters Using Bayesian Optimization for Breast Cancer Prediction
The study presents a method for iterative parameter tuning of tree ensemble-based models using Bayesian hyperparameter tuning for states prediction, using breast cancer as an example. The proposed method utilizes three different datasets, including the Wisconsin Diagnostic Breast Cancer (WDBC) dataset, the Surveillance, Epidemiology, and End Results (SEER) breast cancer dataset, and the Breast Cancer Coimbra dataset (BCCD), and implements tree ensemble-based models, specifically AdaBoost, Gentle-Boost, LogitBoost, Bag, and RUSBoost, for breast cancer prediction. Bayesian optimization was used to tune the hyperparameters of the models iteratively, and the performance of the models was evaluated using several metrics, including accuracy, precision, recall, and f1-score. Our results show that the proposed method significantly improves the performance of tree ensemble-based models, resulting in higher accuracy, precision, recall, and f1-score. Compared to other state-of-the-art models, the proposed method is more efficient. It achieved perfect scores of 100% for Accuracy, Precision, Recall, and F1-Score on the WDBC dataset. On the SEER BC dataset, the method achieved an accuracy of 95.9%, a precision of 97.6%, a recall of 94.2%, and an F1-Score of 95.9%. For the BCCD dataset, the method achieved an accuracy of 94.7%, a precision of 90%, a recall of 100%, and an F1-Score of 94.7%. The outcomes of this study have important implications for medical professionals, as early detection of breast cancer can significantly increase the chances of survival. Overall, this study provides a valuable contribution to the field of breast cancer prediction using machine learning.
- 11 January 2024
- Published by SPIIRAS in Informatics and Automation
- Vol. 23 (1) , 129-168
Open AccessIterative Tuning of Tree-Ensemble-Based Models' parameters Using Bayesian Optimization for Breast Cancer Prediction
The study presents a method for iterative parameter tuning of tree ensemble-based models using Bayesian hyperparameter tuning for states prediction, using breast cancer as an example. The proposed method utilizes three different datasets, including the Wisconsin Diagnostic Breast Cancer (WDBC) dataset, the Surveillance, Epidemiology, and End Results (SEER) breast cancer dataset, and the Breast Cancer Coimbra dataset (BCCD), and implements tree ensemble-based models, specifically AdaBoost, Gentle-Boost, LogitBoost, Bag, and RUSBoost, for breast cancer prediction. Bayesian optimization was used to tune the hyperparameters of the models iteratively, and the performance of the models was evaluated using several metrics, including accuracy, precision, recall, and f1-score. Our results show that the proposed method significantly improves the performance of tree ensemble-based models, resulting in higher accuracy, precision, recall, and f1-score. Compared to other state-of-the-art models, the proposed method is more efficient. It achieved perfect scores of 100% for Accuracy, Precision, Recall, and F1-Score on the WDBC dataset. On the SEER BC dataset, the method achieved an accuracy of 95.9%, a precision of 97.6%, a recall of 94.2%, and an F1-Score of 95.9%. For the BCCD dataset, the method achieved an accuracy of 94.7%, a precision of 90%, a recall of 100%, and an F1-Score of 94.7%. The outcomes of this study have important implications for medical professionals, as early detection of breast cancer can significantly increase the chances of survival. Overall, this study provides a valuable contribution to the field of breast cancer prediction using machine learning.
- Open Access
- 30 April 2023
- Published by Sana'a University in مجلة جامعة صنعاء للعلوم التطبيقية والتكنولوجيا
- Vol. 1 (2)
Breast Cancer-Risk Factors and Prediction Using Machine-Learning Algorithms and Data Source: A Review of Literature
Breast cancer (BC) is a major health concern worldwide. It is a complex and multifactorial disease, and identifying its risk factors is crucial for early detection and effective treatment. This review article provides an overview of the literature on breast cancer risk factors, data sources, and machine learning algorithms for prediction. The paper discusses the various risk factors associated with breast cancer, including age, family history, lifestyle choices, and environmental factors. Additionally, the article explores the different data sources used in breast cancer research, including clinical data, genomic data, and lifestyle data. The paper then reviews the different machine-learning algorithms used for breast cancer prediction, including supervised and unsupervised learning. The performance of ML algorithms in predicting BC risk using different data sources was assessed. This review provides valuable insights into the current state of research on BC risk factors and prediction using ML algorithms and data sources, and the findings will be useful for healthcare professionals and researchers working in the field of BC.
- 30 April 2023
- Published by Sana'a University in مجلة جامعة صنعاء للعلوم التطبيقية والتكنولوجيا
- Vol. 1 (2)
Open AccessBreast Cancer-Risk Factors and Prediction Using Machine-Learning Algorithms and Data Source: A Review of Literature
Breast cancer (BC) is a major health concern worldwide. It is a complex and multifactorial disease, and identifying its risk factors is crucial for early detection and effective treatment. This review article provides an overview of the literature on breast cancer risk factors, data sources, and machine learning algorithms for prediction. The paper discusses the various risk factors associated with breast cancer, including age, family history, lifestyle choices, and environmental factors. Additionally, the article explores the different data sources used in breast cancer research, including clinical data, genomic data, and lifestyle data. The paper then reviews the different machine-learning algorithms used for breast cancer prediction, including supervised and unsupervised learning. The performance of ML algorithms in predicting BC risk using different data sources was assessed. This review provides valuable insights into the current state of research on BC risk factors and prediction using ML algorithms and data sources, and the findings will be useful for healthcare professionals and researchers working in the field of BC.