Haya Aldossary

Haya Aldossary

Auto-generated profile, based on publicly available information

Haya Aldossary

Computer Science Department, College of Science and Humanities, Imam Abdulrahman Bin Faisal University, Jubail, Saudi Arabia

Research Focus & Metrics

Academic Career Length

8Years (2019-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
23 July 2025 in SLAS Technology
Verified
Journal Article
Open access
JYJing Yang
Jing Yang
MSMuhammad Abubakar Siddique
Muhammad Abubakar Siddique
HUHafeez Ullah
Hafeez Ullah
GGGhulam Gilanie
Ghulam Gilanie
Lip Yee PorLip Yee Por
Lip Yee Por
TGThippa Reddy Gadekallu
Thippa Reddy Gadekallu

18 February 2025 in IEEE Transactions on Consumer Electronics
Verified
Journal Article
Closed access
Hao TangHao Tang
Hao Tang
ZWZifeng Wang
Zifeng Wang
YZYonghui Zhang
Yonghui Zhang
BXBo Xu
Bo Xu
UBUzair Bhatti
Uzair Bhatti
HAHaya Aldossary
Haya Aldossary
Factory aquaculture has emerged as a pivotal approach for farmers and researchers dedicated to sustainable, low-carbon agricultural practices. With the advent of agricultural consumer electronics, such as automated feeders, sensors, and controllers, integrating sustainable computing into these devices is crucial for optimizing resource usage. This paper introduces the Sustainable Computing-Based Multi-Objective Factory Aquaculture Feeding Scheduling problem (MOFAFS). This framework is engineered to automate feeding processes with advanced consumer electronics while simultaneously minimizing energy consumption and reducing feeding time in dynamic aquaculture environments. MOFAFS needed to address three core challenges: sequencing feeding in rearing ponds, distributing automated feeding vehicles, and selecting the most efficient feed supply station. The proposed solution utilizes the Enhanced Non-dominated Sorting Genetic Algorithm (ENSGA-II), which dynamically adapts computational data through variable-length coding and implements a greedy neighborhood search strategy tailored to the problem’s characteristics. Additionally, a weighted fitness approach is employed to refine the selection of Pareto-optimal solutions, determining the best compromise solution. The validity of the model is confirmed through a case study that demonstrates its capability to optimize feeding schedules and reduce the loads on feeding vehicles. Benchmark tests demonstrate that ENSGA-II not only reduces feeding time and operates consistently, but also significantly outperforms PSO, GWO, WOA, and NSGA-II in optimizing maximum operating load.

3 November 2025 in IEEE Transactions on Consumer Electronics
Verified
Journal Article
Closed access
JYJing Yang
Jing Yang
VGVijay Govindarajan
Vijay Govindarajan
MAMohammad Yahya H. Al-Shamri
Mohammad Yahya H. Al-Shamri
HAHaya Aldossary
Haya Aldossary
AKAmel Ksibi
Amel Ksibi
KQKun Qi
Kun Qi
The rapid expansion of consumer IoT systems, from smart homes to wearables, generates massive heterogeneous data streams while exposing new attack surfaces to AI-enabled threats. Traditional anomaly detection remains typically limited by unimodal learning, static models, centralized processing, and poor adaptability to concept drift. To overcome these challenges, we propose NeuroAgent-X, a self-evolving cognitive agent that integrates neuro-symbolic reasoning, attention-guided multi-modal fusion, and federated swarm intelligence. The framework features a dual-pathway encoder for structured and unstructured data, a reinforcement-driven self-evolution module with episodic and semantic memory, and selective parameter sharing for privacy-preserving distributed learning. Evaluations on SWaT, UCI HAR, and WADI datasets show superior anomaly detection (94.2% precision, 92.7% recall, 93.4% F1, 0.979 AUC) with sub-120ms latency across cloud, fog, and edge deployments. NeuroAgent-X also demonstrates resilience to concept drift (94% retention), zero-day attacks (89.2% F1), and adversarial perturbations (86.5% F1 under ϵ=0.1), outperforming state-of-the-art baselines. Furthermore, its federated design reduces communication overhead by 78% while preserving data locality. These results establish NeuroAgent-X as a scalable and robust paradigm for securing consumer IoT against adaptive, AI-enabled threats.

6 June 2025 in IEEE Transactions on Consumer Electronics
Verified
Journal Article
Closed access
FUFarhan Ullah
Farhan Ullah
IUIrfan Ullah
Irfan Ullah
KKkhalil khan
khalil khan
QWQuan Wang
Quan Wang
SAShabbab Ali Algamdi
Shabbab Ali Algamdi
Giovanni PauGiovanni Pau
Giovanni Pau
Hyperspectral images (HSIs) are highly complex, containing an enhanced spectral dimension compared to conventional images. Deep learning methods are increasingly being applied to process this three-dimensional data for hyperspectral image classification (HSIC). Existing convolutional and transformer-based methods often struggle with capturing fine-grained spectral-spatial dependencies, achieving high accuracy while balancing computational complexity. In order to solve these problems, we propose a novel SXSFormer network for HSIC. This novel approach integrates Squeeze and Expansion (SX) blocks into the Swin Transformer architecture to enhance feature extraction and attention mechanisms in the model. At its core, the novel SX Block recalibrates channel-wise features by temporarily expanding and then compressing channel dimensionality, allowing the model to focus on informative spectral bands and capture complex interdependencies. The SXSFormer, equipped with window-based multi-head self-attention, efficiently captures long-range dependencies with reduced computational complexity by partitioning input data into non-overlapping windows. Additionally, the SX block’s integration within the Swin Transformer enables improved global contextual understanding by selectively weighting the feature maps. We further refine the architecture using various components, such as patch extraction and embedding layers, and a patch merging strategy, ensuring efficient multi-scale feature extraction. Extensive experiments on four benchmark HSI datasets (SA, IP, PU, and KSC) demonstrate that the proposed model achieves remarkable test accuracies of 99.97%, 98.15%, 99.63%, and 98.14%, respectively, outperforming existing state-of-the-art methods. Our proposed approach also shows good generalization ability when applied to new datasets. Overall, our proposed approach represents a promising direction for HSIC.

6 August 2025 in IEEE Transactions on Intelligent Transportation Systems
Verified
Journal Article
Closed access
MKMuhammad Attique Khan
Muhammad Attique Khan
AAAreej Alasiry
Areej Alasiry
MMMehrez Marzougui
Mehrez Marzougui
IBIsa Bayhan
Isa Bayhan
SKSiva Sarana Kuna
Siva Sarana Kuna
HAHaya Aldossary
Haya Aldossary
Integrating Generative AI (GenAI) into Intelligent Transportation Systems (ITS) raises both enormous opportunities and major worries, especially in the areas of privacy and cybersecurity, which are already at the forefront of these developments. Developing and implementing robust security measures to secure sensitive data and address new cyber threats is of utmost importance, especially with the growing dependence on AI technology in transportation networks. This article looks at GenAI and how it may improve ITS intelligence and efficiency while addressing the risks of using it a lot. It delves into the difficulties of protecting AI-driven systems against hostile assaults (AI-MA), particularly emphasizing transportation infrastructure security, intrusion detection, and data privacy. The research stresses the significance of modern encryption methods, real-time monitoring threats, and adaptive security frameworks to ensure ITS are secure and resilient. In addition, it delves into how transportation systems are affected by ever-changing cyber threats, offering proactive security solutions to combat these dangers and strengthen ITS. This paper’s overarching goal is to lay out a course of action for integrating GenAI into ITS in a way that strikes a good balance between fostering innovation and ensuring privacy and security via thorough analysis. The proposed AI-MA model achieves a high threat detection accuracy of 96.2%, a privacy protection score of 91.8%, a computational efficiency of 92.9%, a resilience score of 97.8%, and a network reliability ratio of 92.6% compared to other existing models.

4 August 2025 in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Verified
Journal Article
Open access
FUFarhan Ullah
Farhan Ullah
IUIrfan Ullah
Irfan Ullah
KKkhalil khan
khalil khan
SKSalabat Khan
Salabat Khan
QWQuan Wang
Quan Wang
HAHaya Aldossary
Haya Aldossary
Hyperspectral images are highly complex, containing a richer spectral dimension compared to conventional images. Deep learning methods are increasingly being applied to process this three-dimensional data for hyperspectral image classification (HSIC). The vision transformer model is steadily gaining prominence in computer vision, emerging as a potential alternative to traditional convolutional neural network (CNN) architectures. Although transformers are powerful due to their self-attention mechanisms, they face challenges related to scalability and efficiency, particularly when processing high-resolution images. In order to solve these problems, in this article, we propose a novel Squeeze-SwinFormer network for HSIC. This novel approach integrates spectral squeeze and excitation (SE) blocks into the shifted window (Swin) transformer architecture to enhance feature extraction and attention mechanisms in the model. The SE block dynamically recalibrates channel-wise feature responses, improving the model’s focus on significant features. The Squeeze-SwinFormer, equipped with window-based multihead self-attention, efficiently captures long-range dependencies with reduced computational complexity by partitioning input data into nonoverlapping windows. In addition, the SE block’s integration within the Swin transformer enables improved global contextual understanding by selectively weighting the feature maps. We further refine the architecture using various components, such as patch extraction and embedding layers, and a patch merging strategy, ensuring efficient multiscale feature extraction. The proposed architecture demonstrates superior performance in tasks requiring attention to intricate details in HSIC. The comprehensive experimental results demonstrate that our proposed model achieved higher test accuracies against the state-of-the-art method, with results of 99.93%, 98.31%, 99.50%, and 97.75% on the Salinas, Indian Pines, University of Pavia, and Kennedy Space Center public benchmark HSIC datasets, respectively. Our approach also shows good generalization ability when applied to new datasets. Overall, our proposed approach represents a promising direction for HSIC.