جامعة المعرفة
المصدر المعتمد للمعلومات البحثية والمخرجات هو قاعدة بيانات سكوبس (Elsevier Scopus) مع تدقيق بيانات الباحثين من عمادة البحث العلمي.

Research outputs

المخرجات البحثية

مقالات ومراجعات وأوراق مؤتمرات وفصول كتب، مفهرسة برقم DOI ومربوطة بباحثي الجامعة.

141 مُخرَج نتائج البحث عن «Algorithm»

مقالوصول مغلق202414استشهاد· Semantic Scholar

Assessing AI literacy and attitudes among medical students: implications for integration into healthcare practice

Journal of Health Organization and Management · مجلد 40 (6) · 1065-1082

PURPOSE: This study aims to assess AI literacy and attitudes among medical students and explore their implications for integrating AI into healthcare practice. DESIGN/METHODOLOGY/APPROACH: A quantitative research design was employed to comprehensively evaluate AI literacy and attitudes among 374 Lusaka Apex Medical University medical students. Data were collected from April 3, 2024, to April 30, 2024, using a closed-ended questionnaire. The questionnaire covered various aspects of AI literacy, perceived benefits of AI in healthcare, strategies for staying informed about AI, relevant AI applications for future practice, concerns related to AI algorithm training and AI-based chatbots in healthcare. FINDINGS: The study revealed varying levels of AI literacy among medical students with a basic understanding of AI principles. Perceptions regarding AI's role in healthcare varied, with recognition of key benefits such as improved diagnosis accuracy and enhanced treatment planning. Students relied predominantly on online resources to stay informed about AI. Concerns included bias reinforcement, data privacy and over-reliance on technology. ORIGINALITY/VALUE: This study contributes original insights into medical students' AI literacy and attitudes, highlighting the need for targeted educational interventions and ethical considerations in AI integration within medical education and practice.

AI literacymedical studentsattitudes toward AIartificial intelligence in healthcare
مقالوصول مفتوح202414استشهاد· Semantic Scholar

RETRACTED ARTICLE: Intelligent dynamic spectrum access using fuzzy logic in cognitive radio networks

Discover Applied Sciences · مجلد 6 (1)

Abstract Cognitive radio networks have emerged as a promising solution to the growing problem of spectrum scarcity and inefficient spectrum utilization. By enabling secondary users to opportunistically access underutilized frequency bands while avoiding harmful interference to primary users, cognitive radio networks offer the potential for more efficient and dynamic spectrum allocation. In this paper, we propose an intelligent approach to dynamic spectrum access using fuzzy logic in cognitive radio networks. The key challenge in cognitive radio networks lies in spectrum sensing, decision-making, and channel selection, where uncertainties and variations in the radio frequency spectrum are prevalent. To address these challenges, we employ fuzzy logic as a powerful tool to handle the inherent imprecision and ambiguity in the decision-making process. Fuzzy-based spectrum sensing algorithms allow for robust detection of spectrum opportunities by considering the “fuzziness” in the received signal strength and noise conditions.

cognitive radio networksdynamic spectrum accessfuzzy logicspectrum sensing
مقالوصول مفتوح202413استشهاد· Semantic Scholar

Heat recovery integration in a hybrid geothermal-based system producing power and heating using machine learning approach to maximize outputs

Case Studies in Thermal Engineering · مجلد 63 · 105210-105210

This study offers an in-depth thermodynamic analysis and optimization of an integrated renewable energy system that merges a double-flash geothermal system with a transcritical carbon dioxide Rankine cycle, utilizing machine learning algorithms. The innovative design aims to maximize the concurrent generation of heat and electricity, ultimately benefiting environmental sustainability and energy security. By employing regression machine learning algorithms, the research evaluates and enhances system performance, achieving remarkable R-squared accuracy levels of 98.86 % for heating output and 99.89 % for power output predictions. The thermodynamic modeling, which has been validated against recognized benchmarks, confirms the accuracy of the system's design. Optimization findings indicate that operating pressures between 840 and 870 kPa and pressure ratios of 1.56–1.60 deliver optimal outputs, with power production between 2582 and 2585 kW and heating output ranging from 12260 to 12280 kW. The system reaches its maximum performance at a pressure of 850 kPa and a pressure ratio of 1.57, resulting in a power output of 2583.97 kW and a heating output of 12279.3 kW. These results highlight the potential of combining advanced thermodynamic systems with machine learning methodologies to improve the efficiency and effectiveness of renewable energy sources.

heat recoverycombined heat and powermachine learning optimizationoperating pressure optimization
مقالوصول مفتوح202412استشهاد· Semantic Scholar

Bagging Vs. Boosting in Ensemble Machine Learning? An Integrated Application to Fraud Risk Analysis in the Insurance Sector

Applied Artificial Intelligence · مجلد 38 (1)

Addressing the pressing challenge of insurance fraud, which significantly impacts financial losses and trust within the insurance industry, this study introduces an innovative automated detection system utilizing ensemble machine learning (EML) algorithms. The approach encompasses four strategic phases: 1) Tackling data imbalance through diverse re-sampling methods (Over-sampling, Under-sampling, and Hybrid); 2) Optimizing feature selection (Filtering, Wrapping, and Embedding) to enhance model accuracy; 3) employing binary classification techniques (Bagging and Boosting) for effective fraud identification; and 4) applying explanatory model analysis (Shapley Additive Explanations, Break-down plot, and variable-importance Measure) to evaluate the influence of individual features on model performance. Our comprehensive analysis reveals that while not every re-sampling technique improves model performance, all feature selection methods markedly bolster predictive accuracy. Notably, the combination of the Gradient Boosting Machine (GBM) algorithm with NCR re-sampling and GBMVI feature selection emerges as the most effective configuration, offering superior fraud detection capabilities. This study not only advances the theoretical framework for combating insurance fraud through AI but also provides a practical blueprint for insurance companies aiming to incorporate advanced AI strategies into their fraud detection arsenals, thereby mitigating financial risks and fostering trust systems.

insurance fraud detectionensemble learningbaggingboosting
مقالوصول مغلق202411استشهاد· Semantic Scholar

Quantum Target Recognition Enhancement Algorithm for UAV Consumer Applications

IEEE Transactions on Consumer Electronics · مجلد 70 (3) · 5553-5560

In UAV Consumer Applications, the challenges and methods of current unmanned aerial vehicle (UAV) radar detection technology are examined. The quantum multi-pattern recognition network model and algorithm are analyzed, and the Quantum Multi-Pattern Recognition Algorithm based on Phase Rotation (PRQMPRA) is proposed according to Grover’s algorithm optimization theory. The issue in the Redundancy Quantum Multi-Pattern Recognition Algorithm (RQMPRA), where a decrease in the probability of successful search can be caused by two phase rotations of$\pi $each, is addressed by the optimization algorithm. The pattern recognition capabilities of Error Backpropagation Algorithm (EBPA), the Deep Autoencoder Learning Algorithm based on Cross-Entropy Function (CDAA), RQMPRA, and PRQMPRA are examined using three different datasets. The results indicate that a higher recognition rate and relatively faster processing speed are exhibited by PRQMPRA when error constraints are specified. To study the target detection problem in UAV consumer applications using a pattern classification approach, a radar target detection method based on the Quantum Multi-Pattern Recognition Algorithm is proposed. Experiments for UAV target detection is conducted with the four algorithms, and the research demonstrates that higher detection accuracy and a high discovery probability can be maintained in low signal-to-noise ratio conditions by PRQMPRA.

Grover's algorithmphase rotationlow signal-to-noise ratiobackpropagation
مقالوصول مفتوح202411استشهاد· Semantic Scholar

An Efficient Authentication and Access Control Protocol for Securing UAV Networks Against Anomaly-Based Intrusion

IEEE Access · مجلد 12 · 62750-62764

UAV networks have gained widespread recognition across multiple industries due to their remarkable adaptability, prompting a fundamental transformation in operational procedures. However, utilizing the open wireless communication channel has consistently brought about significant privacy concerns as a prominent aspect within these networks. Moreover, UAVs are typically equipped with limited computing capabilities that hinder their ability to execute complex cryptographic algorithms. In light of these considerations, this paper proposes a security protocol for UAV networks that uses an authentication and access control mechanism to eliminate the possibility of any security breaches. The proposed scheme is based on hyperelliptic curve cryptography (HECC), which employs a smaller key size of 80 bits instead of the 160 bits required by elliptic curve cryptography (ECC). Remarkably, HECC offers equivalent security to other methods such as RSA, ECC, bilinear pairing, etc., and is, therefore, suitable for UAV networks. The proposed protocol is evaluated for security using the well-known Real-Or-Random (ROR) Oracle model. The automated validation of internet security protocols and applications (AVISPA) tool is employed to illustrate the security of the proposed scheme against adversarial scenarios in the on-the-fly model-checker (OFMC) and constraint-logic-based attack searcher (Cl-AtSe) models. Furthermore, the informal security analysis guarantees our scheme’s robustness against possible attacks based on the Canetti-Krawczyk (CK) and Dolev-yao (DY) adversarial models. The comparative analysis of the proposed protocol’s performance with other existing methods demonstrates its strength in terms of computation and communication costs.

UAV networksauthentication protocolsaccess controlhyperelliptic curve cryptography
مقالوصول مفتوح202411استشهاد· Semantic Scholar

IR-QLA: Machine Learning-Based Q-Learning Algorithm Optimization for UAVs Faster Trajectory Planning by Instructed- Reinforcement Learning

IEEE Access · مجلد 12 · 91300-91315

Trajectory planning for unmanned aerial vehicles (UAVs) presents a formidable challenge, particularly in environments where details about the terrain and objectives are not known beforehand. This uncertainty has propelled the adoption of UAVs across diverse sectors such as logistics, weather forecasting, security surveillance, and autonomous driving, where their ability to navigate autonomously is crucial. Recent advancements have leveraged deep learning for robust object detection and tracking, significantly enhancing data acquisition from aerial imagery and UAV-based applications. Despite numerous innovations, the critical issue of optimizing UAV trajectories in unknown environments remains largely unaddressed. This paper introduces a novel approach, the Instructed Reinforcement Q-Learning Algorithm (IR-QLA), which utilizes Received Signal Strength (RSS) as a dynamic reward metric to refine trajectory planning continuously. By integrating a unique instructed reinforcement principle, this method significantly accelerates the learning process, improving the UAVs’ navigational efficiencies. Our findings demonstrate that IR-QLA not only surpasses traditional algorithms like Ant Colony and Particle Swarm Optimization in terms of convergence speed but also minimizes the path length and iteration counts needed for trajectory optimization. Future studies will explore alternative reward sources and advanced algorithms to further enhance the precision and speed of UAV trajectory planning, aiming to establish new benchmarks in the field.

UAV trajectory planningQ-learningreceived signal strengthreward function design
مقالوصول مفتوح202410استشهاد· Semantic Scholar

Game theory-based optimization for efficient IoT task offloading in 6G network base stations

Measurement Sensors · مجلد 33 · 101184-101184

This study aims to optimise computing for intricate jobs within the overlapping coverage of 6G network base stations. A multi-access edge computing network model is created by solving the issues of task offloading. This model involves many base stations and IoT devices. It takes into account factors such as task delay, energy consumption, societal impacts, and economic incentives. The objective of joint optimization is to maximize the revenues of base stations and the utilities of IoT devices by considering base station price, IoT device base station selection, and job offloading mechanisms. The IoT device base station selection is addressed using a many-to-one matching game model, while pricing and task offloading interactions are managed by a two-stage approach based on Stackelberg game theory. The suggested Optimal Pricing and Best Response Algorithm (OBGT), which is based on Game Theory, successfully reaches equilibrium solutions. It demonstrates rapid convergence in simulations and improves both base station profits and IoT device utility. This study integrates data-driven mobile computing systems assurance for futuristic smart vertical networks, enhancing the efficiency of task offloading and optimization.

mobile edge computingStackelberg gameenergy consumptiontask delay
مقالوصول مفتوح20249استشهاد· Semantic Scholar

Design of Artificial Intelligence Driven Crowd Density Analysis for Sustainable Smart Cities

IEEE Access · مجلد 12 · 121983-121993

Smart Cities refer to urban areas which exploits recent technologies for improving the performance, sustainability, and livability of their infrastructure and services. Crowd Density Analysis (CDA), a vital component of Smart Cities, involves the use of sensors, cameras, and data analytics to monitor and analyze the density and movement of people in public spaces. CDA utilizing DL harnesses the control of neural networks to mechanically and exactly evaluate the density of crowds in numerous settings, mainly in smart cities. DL techniques like Recurrent Neural Network (RNN) and Convolutional Neural Network (CNN), are trained on vast datasets of crowd videos or images to learn complex designs and features. These models can forecast crowd density levels, recognize crowd anomalies, and offer real-time visions into crowd behavior. This study designs an Artificial Intelligence Driven Crowd Density Analysis for Sustainable Smart Cities (AICDA-SSC) technique. The aim of the AICDA-SSC method is to analyze the crowd density and classify it into multiple classes by the use of hyperparameter-tuned DL models. To accomplish this, the AICDA-SSC technique applies contrast enhancement using the CLAHE approach. Besides, the complex and intrinsic features can be derived by the use of the Inception v3 model and its hyperparameters can be chosen by the use of the marine predator’s algorithm (MPA). For crowd density detection and classification, the AICDA-SSC technique applies a gated recurrent unit (GRU) model. Finally, a chaotic sooty tern optimizer algorithm (CSTOA) based hyperparameter selection procedure takes place to increase the effectiveness of the GRU system. The experimental evaluation of the AICDA-SSC technique takes place on a crowd-density image dataset. The experimentation values showcase the superior performance of the AICDA-SSC method to the recently developed DL models.

smart citiesgated recurrent unitmarine predators algorithmhyperparameter optimization
مقالوصول مفتوح20248استشهاد· Semantic Scholar

Blockchain-based secure optimized traceable scheme for smart and sustainable food supply chain

Discover Sustainability · مجلد 5 (1)

Abstract A large number of nodes, and significant system overhead are among the challenges. To tackle these problems, a new PBFT technique known as trace-PBFT (t-PBFT) is being contemplated for adoption in the food supply chain. The PBFT algorithm, short for Practical Byzantine Fault Tolerance, serves as the fundamental algorithm. The linkages throughout the supply chain may be categorized into three primary groups. As a response to the continuing discussion over the amount of data, the status of each node is continuously updated in real time. This metric is utilized to evaluate the dependability of the principal node and to analyze the overall performance of the nodes. To minimize the amount of data transmitted between nodes, we optimized the consensus process of the first algorithm. The distinct attributes of offering collaborative food technology were considered throughout this procedure. Experimental evidence indicates that the t-PBFT approach outperforms the PBFT algorithm in terms of throughput, request delay, and information overhead. An architectural paradigm is proposed by integrating the t-PBFT algorithm with a federated network. This model was developed in response to the distinct demands of the food supply chain. This model effectively captures data at each stage of the food supply chain to ensure information tracking while preserving the safety of the food technology flow process.

blockchainfood supply chainPBFTconsensus optimization
مقالمفتوح (مستودع)20247استشهاد· Semantic Scholar

Performance Assessment of CPU Scheduling Algorithms: A Scenario-Based Approach with FCFS, RR, and SJF

Journal of Computer Science · مجلد 20 (9) · 972-985

This study presents an extensive examination of CPU scheduling algorithms, focusing on the First-Come, First-Served (FCFS), Round-Robin (RR), and Shortest-Job-First (SJF) strategies through a carefully designed scenario-based approach. By deploying a Java-based simulation to dynamically generate random process arrival and burst times, this study simulates a variety of operational conditions to test these scheduling algorithms’ adaptability and performance in environments that closely resemble real-world computing scenarios. The research aims to explore the effects of dynamic quantum size allocation on RR scheduling and assess its impact on system performance metrics such as response time and context switching overhead. Through a detailed analysis, this study seeks to provide new insights into the operational efficiency of the FCFS, RR, and SJF scheduling strategies, highlighting their strengths, limitations, and applicability across different computing environments.

process schedulingfirst come first servedround robin schedulingshortest job first scheduling
مقالوصول مفتوح20247استشهاد· Semantic Scholar

Meta-Learning Guided Weight Optimization for Enhanced Solar Radiation Forecasting and Sustainable Energy Management with VotingRegressor

Sustainability · مجلد 16 (13) · 5505-5505

Solar radiation prediction plays a crucial role in renewable energy management, impacting various decision-making processes aimed at optimizing the utilization of solar resources and promoting sustainability. Ensemble regression methods, notably VotingRegressor, have emerged as promising tools for accurate solar radiation forecasting. By integrating predictions from multiple base estimators, ensemble methods have the potential to capture intricate patterns inherent in solar radiation data. However, achieving optimal predictive performance with ensemble methods heavily relies on the careful weighting assigned to each base estimator, presenting a significant challenge. In this study, a novel approach is presented to enhance solar radiation prediction by utilizing meta-learning techniques to optimize the weighting mechanism in the VotingRegressor ensemble. Meta-learning, a subfield of machine learning focusing on learning algorithms across different tasks, provides a systematic framework for learning to learn. This enables models to adapt and generalize more effectively to new datasets and tasks. Our proposed methodology demonstrated significant improvements, with the VotingRegressor with meta-learning techniques achieving an RMSE of 8.7343, an MAE of 5.42145, and an R² of 0.991913. These results mitigate the need for manual weight tuning and improve the adaptability of the VotingRegressor to varying solar radiation conditions, ultimately contributing to the sustainability of renewable energy systems. The methodology involves a comprehensive exploration of meta-learning techniques, encompassing gradient-based optimization, reinforcement learning, and Bayesian optimization.

solar irradiance forecastingmeta-learningensemble regressionweight optimization