A Study of Some Neutrosophic Derivatives Problems Based On Newton's BDF and CDF Numerical Methods

Ahmad A.; Mayada; Ahmed Atallah; Abdallah

doi:https://doi.org/10.54216/IJNS.260401

Journal of Cybersecurity and Information Management | 2026

An Empirical Evaluation of the Main Factors of a Cybersecurity Culture in South African E-health Institutions Using Multiple Linear Regression

E-health institutions are prominent targets for cybercriminals due to their reliance on information technology systems and issues related to the users have been identified as the biggest security weakest. Hence, while cybersecurity culture (CSC) research emphasizes the necessity of the human factor, limited empirical work has been done in the context of e-health in Africa. Therefore, an empirical evaluation was conducted to identify how preparedness, responsibility, management, technology and environment influence cybersecurity in South African e-health institutions. This quantitative research studied e-health institutions in the Mpumalanga province of South Africa. Various methods were used to investigate the multiple linear regression effects of the main factors of CSC and the results show that although the preparedness (Beta = 0.281; p-value < 0.05) and environment (Beta = 0.500; p-value < 0.05) factors had the greatest influence, management, technology and environment had a positive effect on CSC. These factors contributed 48.2 % to the variance (R-Squared). The study seems to be the first empirical study that combines the human factor domain framework (HFD) with other theoretical frameworks to identify critical factors of CSC. Furthermore, the impact of technology on CSC was empirically tested. The study is significant as it identified key factors that contributed to the institution’s CSC and quantified their impact. These results can enable e-health institutions to make decisions based on evidence regarding their cybersecurity interventions, strategy and practices. However, the empirical evaluation was limited to one context, namely the Mpumalanga province in South Africa and at two hospitals selected based on easy access (convenience) and purposive sampling with criteria based on work experience and knowledge of CSC limited the number of participants eligible to participate.

groups

Nwanneka E. Mwim mail -

Jabu Mtsweni mail -

Bester Chimbo mail

link https://doi.org/10.54216/JCIM.170213

Volume & Issue

Vol. Volume 17 / Iss. Issue 2

Details open_in_new

Journal of Intelligent Systems and Internet of Things | 2025

Advanced Predictive Analysis of EGDI Time Series Using Hybrid ARIMA-LSTM and SARIMAX: A Comparative Study for Iraq and Tunisia

This study presents a predictive modeling framework for forecasting the E-Government Development Index (EGDI) using two advanced time series approaches. Firstly, the Seasonal Auto Regressive Integrated Moving Average with Exogenous Variables (SARIMAX). Secondly, hybrid ARIMA-LSTM model. We focus on two case studies, Iraq and Tunisia, based on monthly EGDI data from the United Nations Survey Reports, spanning the years 2003 to 2024. Using several preprocessing steps such as handling missing data, testing for stationarity using the combined ADF and KPSS tests, and determining the optimal ARIMA parameters through ACF and PACF analysis and implementing autoarima. The model was built and trained using 80% of the data, while 20% was retained for testing. The independence of the residuals verified using the Ljung-Box test. Four types of visualization and error analysis were applied using ACF/PACF for residuals, error plots as prediction error plot, error distribution plot (histogram + KDE) and decomposition analysis to visually assess model fit. Evaluation was conducts using multiple error metrics, including RMSE, MAE, MAPE, MHE, AIC, BIC and MAPA. After building the four models, we ensured that the results and reconstructions were evaluated using the 12 tests we mentioned, and that they were based on the best results and were consensus acceptable. ARIMAX model demonstrated superior performance, achieving an average absolute percentage Accuracy (MAPA) of 98.35% for Iraq and 97.93% for Tunisia. In comparison, the hybrid ARIMA-LSTM model, which combines linear ARIMA outputs with nonlinear corrections from an LSTM neural network, demonstrated competitive predictive ability with a MAPA of 95.68% for Iraq and 96.14% for Tunisia.  SARIMAX showed slightly outperformed the hybrid model in overall accuracy. On other hand, ARIMA-LSTM model demonstrated robustness in capturing complex nonlinear dynamics particularly in the more structurally diverse Tunisian dataset. These results confirm the potential of both models as effective tools for predicting EGDIs and support their application in digital governance planning and policymaking. We designed and we recommend adopting our "12 -Test Approach" for evaluation framework as a standard methodology in future studies addressing analysis and forecasting, and its suitability for different types of time series models. This approach provides comprehensiveness, accuracy, and flexibility in evaluation, regardless of model type or application area.

groups

Ali Ahmed Ali mail -

Atef Masmoudi mail

link https://doi.org/10.54216/JISIoT.170206

Volume & Issue

Vol. Volume 17 / Iss. Issue 2

Details open_in_new

Journal of Intelligent Systems and Internet of Things | 2025

Designing Explainable Deep Learning Models for Biomedical Data Analysis and Clinical Prediction Enhancement

Recent advancements in biomedical data analysis have significantly transformed clinical decision-making. However, the inherent complexity and heterogeneity of healthcare data continue to present major challenges. Traditional deep learning models, while powerful, often lack transparency, limiting their adoption in clinical settings due to their "black-box" nature. To address this critical gap, this study introduces a novel Explainable Deep Learning (XDL) framework that integrates high predictive accuracy with interpretability, enabling clinicians to trust and validate AI-driven insights. The proposed framework leverages advanced interpretability techniques—such as Grad-CAM for visual attribution and SHAP for feature importance analysis—to analyze multimodal biomedical data, including clinical imaging, genomic sequencing, and electronic health records. Experimental evaluations across three benchmark datasets demonstrated the model’s strong performance, achieving an accuracy of 91%, sensitivity of 95.4%, specificity of 98.6%, and an AUC of 99%, while maintaining an interpretability score of 92% as rated by domain experts. Compared to non-explainable models, the proposed approach showed a 12.3% increase in interpretability and a 5.8% improvement in accuracy. Importantly, attention map analysis revealed alignment with clinically relevant biomarkers in 93% of cases and uncovered previously overlooked prognostic patterns in 18% of patient cohorts. These findings underscore the model’s potential to enhance diagnostic precision and support more informed clinical decisions. Moreover, the algorithm reduced diagnostic time by 23% due to its provision of actionable insights. The hybrid approach—combining built-in attention mechanisms with external interpretability tools—ensures seamless integration into clinical workflows while supporting compliance with regulatory standards for transparency.

groups

Maha Rahrouh mail -

Walid Alayash mail -

Inas salah Mahmoud mail -

Marwa Hussien Moahmed mail

link https://doi.org/10.54216/JISIoT.170207

Volume & Issue

Vol. Volume 17 / Iss. Issue 2

Details open_in_new

Journal of Intelligent Systems and Internet of Things | 2025

Ensemble Learning-Based Intrusion Detection and Classification for Securing IoT Networks: An Optimized Strategy for Threat Detection and Prevention

The Internet of Things (IoT) advancement has created new security holes, which require intrusion detection systems to defend networks effectively. The complex structure of IoT networks causes traditional security methods to fail because they produce high amounts of incorrect detections and limited ability to accurately identify threats. The authors introduce ID-ELC: Ensemble Learning and Classification framework for Intrusion Detection, which aims to strengthen IoT environment security. A new ID-ELC model uses CS optimization with composite variance to choose network features that boost their detection capabilities. The cybersecurity evaluation of the system utilized Kyoto network records that included 91,000 intrusion-prone records and 59,000 benign logs from 150,000 total records. Experiments revealed ID-ELC surpasses Statistical Flow Features (SFF) and Two-layer Dimension Reduction and Two-tier Classification (TDRTC) through precision 0.98, accuracy 0.98, sensitivity 0.99 and specificity 0.97. Science-based evaluations confirm ID-ELC represents a flexible and resilient tool for IoT intrusion protection that shows practical value for citywide security systems and medicine networks and manufacturing operations. Future investigation will concentrate on enhancing the selection of features alongside classification methods to address rising cyber threats.

groups

Kumaresh Sheelavant mail -

Charan K. V. mail -

B. Yamini Supriya mail -

Purshottam J. Assudani mail -

Chandra Bhushan Mahato mail -

Sanjay Kumar Suman mail

link https://doi.org/10.54216/JISIoT.170208

Volume & Issue

Vol. Volume 17 / Iss. Issue 2

Details open_in_new

Fusion: Practice and Applications | 2025

Deep Learning Approaches for Automated Disease Detection in Agriculture

This research introduces a cutting-edge deep learning-based agricultural engineering illness diagnosis approach. Convolutional neural networks (CNNs) and improved methods improve accuracy and efficiency. The recommended solution includes network settings, convolution processes, and sharing strategies to reduce dimensions. These methods reduce the network's processing power so it can concentrate on disease characteristics. The model employs dropout regularization, attention processes, and multi-scale feature extraction to enhance sickness prediction. The technology also utilizes photographs and sensor data to adapt to agricultural circumstances. The performance test shows that the suggested technique outperforms traditional machine learning and mixed models in F1 score (95%), accuracy (95%), precision (94%), memory (96%), and correctness (94%). It has high discriminative power with an AUC-ROC score of 0.98. The model uses computers well: two hours to train, two seconds to derive conclusions, and 65% of the CPU at all times. Real-time farming could benefit from its use. The suggested technique can properly and reliably diagnose illnesses due to its low overfitting rate and excellent generalization potential. The precision agriculture technique will enhance crop health management and productivity.

groups

Ahmed A. F. Osman mail -

Rajit Nair mail -

Mosleh Hmoud Al-Adhaileh mail -

Theyazn H.H Aldhyani mail -

Saad M. AbdelRahman mail -

Sami A. Morsi mail

link https://doi.org/10.54216/FPA.200204

Volume & Issue

Vol. Volume 20 / Iss. Issue 2

Details open_in_new

Fusion: Practice and Applications | 2025

Personalized Cognitive Behavioral Therapy for Adults Using Machine Learning: A Multi-Factor, Reinforcement-Based Approach

This paper presents a novel machine-learning framework designed to personalize Cognitive Behavioral Therapy (CBT) for adult patients by leveraging a multi-dimensional, adaptive approach. The proposed system integrates historical clinical data, real-time behavioral indicators, and contextual factors to generate a comprehensive psychological profile for each adult patient. A reinforcement learning mechanism underpins therapy selection, allowing the model to iteratively refine treatment strategies based on individual responses and therapeutic outcomes. An embedded optimization process enables dynamic adaptation of interventions, improving predictive accuracy and fostering patient-centered care. The framework incorporates a multi-factor assessment model that synthesizes psychological, behavioral, and physiological variables to enhance therapeutic effectiveness, sustainability, and responsiveness to change. Comparative evaluations demonstrate that this approach outperforms traditional CBT planning methods, as well as existing deep learning, hybrid, and reinforcement-based models, in terms of accuracy, interpretability, computational efficiency, and patient outcome optimization for adults. Furthermore, the system emphasizes fairness and equity in treatment personalization, supporting real-time clinical decision-making while minimizing ineffective therapeutic pathways. This research underscores the transformative potential of machine learning in mental health care by enabling scalable, data-driven, and continuously improving interventions tailored to the nuanced needs of adult patients undergoing CBT.

groups

Mohammed Awad Alasmrai mail -

Ramadan Mohamed Ismail mail -

Mohammed Hasan Ali Al-Abyadh mail

link https://doi.org/10.54216/FPA.200205

Volume & Issue

Vol. Volume 20 / Iss. Issue 2

Details open_in_new

Fusion: Practice and Applications | 2025

Machine Learning for Free Space Optical Communication: A Systematic Review with Emphasis on NOMA and Massive MIMO Integration

Advancements in high-speed communication networks, such as 5G and 6G, display the shortcomings of earlier Radio Frequency (RF) systems due to their limited access to the electromagnetic spectrum. Optical Wireless Communication (OWC) gives access to an unlimited optical spectrum that can address the demands in 6G networks. One key thing about Free Space Optical (FSO) is that it uses the near-infrared spectrum to transfer large amounts of data over several kilometers. FSO systems can be found in a large number of places, ranging from home and outdoor use to important roles in the military and in medical settings. These systems, however, struggle to transmit signals clearly and reliably when the distance is very long due to effects of the atmosphere. One solution to these problems is to rely on advanced channel modeling and using Multiple-Input Multiple-Output (MIMO) schemes, as they improve reliability and efficiency. The latest research efforts are centered on Massive MIMO-FSO networks that make use of spatial diversity to fight atmospheric fading and guarantee a sturdier connection. Importantly, Machine Learning (ML) is transforming the way research is carried out. Channel estimation, turbulence prediction, signal demodulation, and adaptive modulation can now be done using ML, which reduces the need for many calculations and makes things run more smoothly. Using information from data, ML helps optimize FSO systems in different channel conditions. This study provides a review of how machine learning is applied in Massive MIMO-FSO systems. It sorts out highlighting current strategies, explaining their strengths, weaknesses, and how to use them. The main goal of this review is to give an in-depth look at how ML-assisted optical wireless systems can fulfill the needs of future communication networks.

groups

Hasan Farooq Radeef mail -

Lwaa F. Abdulameer mail -

Heba M. Fadhil mail

link https://doi.org/10.54216/FPA.200206

Volume & Issue

Vol. Volume 20 / Iss. Issue 2

Details open_in_new

Fusion: Practice and Applications | 2025

A Two-Stage System for Surveillance Video Summarization and Unsupervised Abnormal Event Detection in Educational Institutions

Surveillance cameras play a pivotal role in educational institutions. They monitor the educational process, detect violations, and protect students from potential injuries or dangers. Continuous recording generates a massive amount of video data. Human observers spend significant time and effort reviewing the footage. Reviewing aims to detect and quickly address abnormal events. Abnormal events are rare in educational environments. Observers may become bored during continuous monitoring. This may cause fatigue and loss of attention. To overcome these challenges, this paper proposes an intelligent system that combines summarization and abnormal event detection in surveillance video. It is divided into two stages: The first stage starts with the extraction of static, feature-based key frames that highlight the video's most significant content. In the second stage, Convolutional Autoencoder (CAE) network used to detect abnormal events from the key frames generated by the summary stage. The proposed system produces two separate videos: a general summary and a dedicated abnormal events video sent to the relevant individuals. The proposed system was tested on some benchmark datasets. The experimental results demonstrated that the proposed system was effective in reducing browsing time and effort, as well as in detecting abnormal events within an educational context.

groups

M. E. ElAlmi mail -

M. M. Lotfy mail -

M. M. Ghoniem mail

link https://doi.org/10.54216/FPA.200207

Volume & Issue

Vol. Volume 20 / Iss. Issue 2

Details open_in_new

Fusion: Practice and Applications | 2025

Generative AI Fusion in Digital Learning: Educators Insights in Revolutionising Modern Education

The advancement and expansion of artificial intelligence (AI) has revolutionized traditional education paradigms. The ability of language models to process human language has revolutionized the field of artificial intelligence. This had led to the integration of language models such as Generative AI (GAI) into learning as it can understand and process human language efficiently. Fusion of these models has significantly enhanced education and research development leading to academic progress. There is gap in the learning needs of the students. Traditional teaching methods often fail to provide personalised adaptive environments and hence to fill this gap this research focusses on integration of AI tools in classrooms.  The objective of this paper is to explore and analyze the applications of integration of generative AI strategies in teaching and to examine the impact from educators’ perspective. The objective of the study is to evaluate the effectiveness of GAI powered integration in teaching and learning by analyzing the feedback scores gathered by students and teachers of an undergraduate course. Data was collected and analyzed using standard mean comparisons. Results of the analysis demonstrate that generative AI assisted teaching facilitated adaptive learning, automated content generation, enhanced student engagement and the quality of dynamic learning when compared with conventional strategies. Using quantitative analysis, the study validates GAI fusion, and the data is analyzed using standard mean scores.  The improvement performance of students and educators feedback for traditional and GAI is 56.63% and 54.41% respectively, which suggests a positive shift of moving from traditional to GAI approaches. This strong score shows the GAI approach is more effective and student-centered. The results reveal that though challenges exist, strategic guided integration of GAI significantly enhances pedagogical factors of education and thus plays a crucial role in shaping AI education as AI models evolve.

groups

Moosa Ahmed Hassan Bait Ali Sulaiman mail -

Anita Venugopal mail

link https://doi.org/10.54216/FPA.200208

Volume & Issue

Vol. Volume 20 / Iss. Issue 2

Details open_in_new

International Journal of Neutrosophic Science | 2025