Duhok Polytechnic University

The increasing complexity of constrained engineering design problems has intensified the demand for metaheuristic optimization algorithms that are both computationally efficient and robust against premature convergence and search stagnation. This paper presents the Signal Lock Optimization Algorithm (SLOA), a novel population-based metaheuristic founded on the dual mechanisms of confidence reinforcement and noise suppression. The core principle of SLOA lies in identifying high-confidence solution components-referred to as signal locks-and reinforcing them during the search process while dynamically filtering stochastic perturbations that may mislead exploration in multimodal and highly constrained landscapes. The proposed algorithm incorporates adaptive parameter updating and an effective constraint-handling strategy to maintain a balanced exploration–exploitation trade-off. SLOA is extensively evaluated on a suite of widely adopted engineering benchmark problems, including Welded Beam Design, Pressure Vessel Design, Tension/Compression Spring Design, and Car Side-Impact Design. Comparative experimental results and statistical analyses demonstrate that SLOA consistently achieves superior or highly competitive solution quality, faster convergence rates, and high feasibility compared to several state-of-the-art metaheuristic algorithms. The findings confirm that the signal lock mechanism provides a reliable and scalable optimization framework for solving complex real-world engineering design problems.

The construction industry is undergoing a transformative shift with the integration of intelligent monitoring technologies. Advanced sensing systems, including IoT-enabled devices, drones, and wearable technologies, are increasingly employed to track project progress, detect potential hazards, and predict environmental impacts. Coupled with artificial intelligence (AI) and machine learning algorithms, these technologies facilitate proactive decision-making, enhance operational efficiency, and reduce human error. This chapter explores emerging trends in intelligent construction monitoring, highlighting innovations in real-time data acquisition, predictive analytics, automated reporting, and adaptive risk management frameworks. It also examines challenges related to data integration, interoperability, cybersecurity, and ethical considerations, providing a comprehensive roadmap for researchers, practitioners, and policymakers to leverage intelligent monitoring systems for safer and more sustainable construction practices.

The rapid digital transformation of the construction industry has driven the adoption of intelligent construction monitoring systems to enhance safety, productivity, and risk management in complex and dynamic project environments. These systems integrate advanced sensing technologies, artificial intelligence, Internet of Things (IoT) platforms, and data analytics to enable real-time observation, prediction, and decision-making throughout the construction lifecycle. This chapter presents a comprehensive overview of the foundations, system architectures, and emerging trends in intelligent construction monitoring systems. It examines key enabling technologies, including computer vision, machine learning, digital twins, and cloud–edge computing frameworks, and discusses how these components are orchestrated to support real-time safety monitoring, environmental impact prediction, and proactive risk mitigation.

Food waste has become a pressing global sustainability challenge, reflecting both environmental degradation and inefficient resource utilization. This study explores the relationships between attitudes, subjective norms, and perceived behavioral control (PBC) toward food waste reduction in the Kurdistan Region of Iraq, guided by the Theory of Planned Behavior (TPB). A quantitative approach was applied using an online survey of 205 respondents from Duhok City, employing a convenience sampling technique. Data were analyzed using SPSS version 26 through descriptive statistics, reliability and validity tests (Cronbach’s alpha, KMO, and Bartlett’s test), and multiple regression analysis. The results revealed that attitudes (? = 0.46, p < .001) and subjective norms (? = 0.30, p < .001) significantly influence perceived behavioral control (R² = 0.314), indicating that moral attitudes and social influence are key predictors of individuals’ perceived capacity to reduce food waste. The study confirms the applicability of the Theory of Planned Behavior in understanding food waste reduction–related perceptions in emerging economies and highlights the importance of public education, social campaigns, and household awareness programs. The findings provide valuable insights for policymakers, educators, and environmental organizations seeking to foster behavioral change and sustainability in the Kurdistan Region of Iraq.

<title>Abstract</title> Continuous Sign Language Recognition (CSLR) seeks to transcribe unsegmented sign language videos into gloss sequences without frame-level supervision, presenting persistent challenges in temporal alignment, long-range dependency modeling, and reliable sequence-level generalization. While recent advances have achieved strong performance in high-resource languages, Kurdish Sign Language (KrdSL) remains largely unexplored due to the absence of sentence-level benchmarks. To address this gap, we introduce <bold>KrdSL-1400</bold> , the first continuous Kurdish Sign Language dataset, comprising 1,400 annotated video sequences covering 40 linguistically structured sentences performed by seven native signers, providing a standardized benchmark for low-resource CSLR. We propose a <bold>hybrid spatio-temporal CSLR framework</bold> that combines deep convolutional visual encoding with sequence-aware temporal modeling and a <bold>multi-task joint CTC–attention decoding strategy</bold> explicitly designed to address alignment uncertainty. The CTC objective enforces monotonic alignment, while an <bold>entropy-regularized multi-head attention mechanism</bold> dynamically emphasizes linguistically salient temporal segments, enabling robust sequence prediction without reliance on pose estimation or handcrafted features. Training dynamics exhibit stable and consistent convergence, with closely aligned training and validation WER curves indicating strong generalization. Quantitative evaluation shows that a CTC baseline achieves a WER of <bold>13.5%</bold> , which is reduced to <bold>10.5%</bold> using single-head attention, while the proposed model attains the best performance with a WER of <bold>9.5%</bold> , corresponding to an approximate <bold>30% relative improvement</bold> . <bold>Cross-dataset evaluation</bold> on the large-scale <bold>PHOENIX-2014-T</bold> benchmark further demonstrates generalization, achieving a WER of <bold>13.7%</bold> and outperforming recent attention-based and transformer-based CSLR approaches.

Earlier diagnosis of pulmonary disease is greatly significant in enhancing treatment impacts and lowering systems' medical workload. As increasingly more cases of pulmonary disease accumulate, methods of deep learning (DL) have increasingly become a viable option towards assisting physicians with diagnoses, particularly through the interpretation of chest X-ray (CXR) images. This paper presents the latest DL-based models and methods for early detection of lung diseases and evaluates their performance and accuracy of disease classification. It also demonstrates the power of ensemble learning methods, a combination of ResNet, EfficientNet, and Inception models, for enhancing the accuracy and reliability of diagnosis systems, especially in handling complicated patterns of diseases. The study seeks to introduce newer directions for research and explore the direction towards intelligent and scalable diagnosis solutions that can potentially make a critical contribution towards enhanced early detection and improving the quality of patient care.

Recent advances in deep learning have significantly transformed medical diagnostics, particularly in dermatology. Accurate skin disease detection and classification areessential for effective treatment and improved patient outcomes. This systematic review examines deep learning approaches, including Convolutional Neural Networks (CNNs) and transfer learning, for automated dermatological diagnosis. Public datasets such as HAM10000 and ISIC play a key role in training robust models; however, challenges including dataset imbalance, disease heterogeneity, and overfitting remain. Techniques such as ensemble learning, attention mechanisms, explainable artificial intelligence, data augmentation, hybrid models, and task-specific loss functions have been shown to enhance accuracy, robustness, and interpretability. This study follows a systematic review methodology in accordance with the PRISMA guidelines. The review synthesizes 17 studies published between 2021 and 2024, highlighting the potential of deep learning to support scalable and reliable dermatological diagnostic systems.

Cloud computing, web technology, and digital marketing are converging together to change enterprise systems into more intelligent, more responsive platforms. AI improves decision-making and personalization, web interfaces facilitate real-time interaction, and cloud infrastructure provides scale. All these advances notwithstanding, there remain challenges with data privacy, bias in algorithms, and complexities of implementation. The objective of this paper is to analyze how these technologies, when brought together, enhance the performance of enterprises alongside their limitations and present methods for creating intelligent, sustainable, and ethics-minded enterprise solutions.

Biomedical question answering requires accurate and interpretable systems; however, existing approaches often face challenges such as language model hallucinations and limited reasoning when relying solely on standalone knowledge graphs. To address these limitations, this study proposes a hybrid framework that integrates the LLaMA-3B language model with a Neo4j-based drug–disease–symptom knowledge graph. The system translates natural language questions into executable Cypher queries, operates on an iBKH-derived graph comprising over 65,000 entities and 3 million relationships, and returns answers with supporting evidence through a transparent interface. Experiments conducted on 60 biomedical questions across three levels of difficulty demonstrate the robustness of the approach: 96% exact match for simple queries, 95% for medium queries, and 86.7% for complex queries. Overall, the system achieves Precision@5 of 96.1%, Recall@5 of 89.0%, F1@5 of 91.0%, Hits@k of 96.1%, and an MRR of 94.4%, while maintaining an average response time of only 6.07 seconds. These results indicate that the system retrieves nearly all relevant answers, ranks them correctly, and delivers them with latency low enough for interactive use. Moreover, unlike cloud-based APIs such as ChatGPT, which require internet connectivity and external data transmission, the proposed framework operates fully offline, ensuring privacy, reproducibility, and compliance with biomedical data governance. Overall, this pipeline provides an accurate, efficient, and privacy-preserving solution for biomedical question answering, making it a practical alternative to cloud-dependent approaches in sensitive healthcare contexts.

Breast cancer is an alarming worldwide health concern, and early detection is crucial for improving patient outcomes. This study explores the application of deep learning algorithms in breast thermography, a non-invasive and radiation-free imaging tech-nique, to enhance diagnostic accuracy. This research synthesizes peer-reviewed lit-erature from 2020 to 2025, focusing on various deep learning architectures, including CNNs, GANs, RNNs, U-Net, and transfer learning, in relation to thermographic da-tasets like DMR-IR and Visual DMR. The methodology employs a structured approach encompassing literature searches, criteria for inclusion and exclusion, data extraction, and synthesis. The findings indicate that deep learning significantly enhances seg-mentation, classification, and anomaly detection in thermal breast images, frequently surpassing traditional diagnostic techniques. While accuracy rates are promising, challenges persist, such as limitations in datasets, variability in images, and a lack of standardization. This study highlights the potential of AI-enhanced thermography as a cost-effective and scalable method for breast cancer screening, while also identifying key areas for further research to enhance generalizability and clinical application.