Egyptian Informatics Journal最新文献

{"title":"Enhanced Intrusion Detection System Using Hybrid-Inspired Algorithms and Conditional Generative Adversarial Networks for Internet of Things Security","authors":"Shahab Wahhab Kareem","doi":"10.1016/j.eij.2025.100763","DOIUrl":"10.1016/j.eij.2025.100763","url":null,"abstract":"<div><div>This research proposes a new Intrusion Detection System architecture that aims at improving the protection of IoT through the integration of multiple techniques. The system utilizes a dataset called Bot-IoT that contains unbalanced data, to develop a stable Intrusion Detection System. The methodology is divided into three stages: data pre-processing, synthetic data generation and bio-inspired hybrid features selection. The first process includes cleaning, encoding as well as scaling the data set for the machine learning algorithms such as Extreme Gradient Boosting (XGBoost), Light Gradient Boosting Machine, Deep Neural Network, Support Vector Machine, and Decision Tree to be affected. The second stage involves the use of a novel architecture of a Conditional Generative Adversarial Network, with a two-discriminator structure to enhance the generation of synthetic data that will improve the balance and the overall quality of the dataset used for training. It is described how the validity of the synthetic data is assessed statistically, and how the generated data affects different models of machine learning. The last step uses <em>meta</em>-heuristic bio-hybrid algorithms for selecting features. The two methods of crocodile hunting search and bee optimization are integrated with Recursive Feature Elimination to select superior features from the dataset used in the experiment. The integration of these models allows for achieving the highest levels of detection with a minimum of false positives. The research advances the field of artificial intelligence by enhancing Conditional Generative Adversarial Networks (CGAN) with a two-discriminator architecture for synthetic data generation, coupled with a novel hybrid feature selection algorithm. This AI innovation is applied to the development of an Intrusion Detection System (IDS) aimed at improving the cybersecurity of Internet of Things (IoT) networks.“</div></div>","PeriodicalId":56010,"journal":{"name":"Egyptian Informatics Journal","volume":"31 ","pages":"Article 100763"},"PeriodicalIF":4.3,"publicationDate":"2025-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145019087","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Symbolic regression and interpretable ensemble learning approach in determining early onset of diabetic peripheral neuropathy","authors":"Abhigyan Nath ,&nbsp;Rachita Nanda ,&nbsp;Prajna Parimita Jena ,&nbsp;Amritava Ghosh ,&nbsp;Seema Shah ,&nbsp;Suprava Patel ,&nbsp;Eli Mohapatra ,&nbsp;Anoop Kumar Tiwari ,&nbsp;Kottakkaran Sooppy Nisar","doi":"10.1016/j.eij.2025.100777","DOIUrl":"10.1016/j.eij.2025.100777","url":null,"abstract":"<div><div>The detrimental consequences of diabetic peripheral neuropathy (DPN), a prevalent comorbidity of type 2 diabetic mellitus, include heightened morbidity and mortality, as well as a reduced quality of life. Unfortunately, this condition has been identified on a frequent basis in recent years, but it is either inadequately diagnosed or remains untreated. Diabetic peripheral neuropathy is caused and advances by a complex interplay of metabolic process imbalance, immune system dysfunction, oxidative stress, and endothelial dysfunction each of which has an impact on its multifactorial pathogenesis. Acquiring information on the physiological traits associated with the DPN group can assist in recognizing and explaining the possible development of early warning systems. The key goal of this investigation is to comprehensively examine the physiological traits that differentiate the DPN and No DPN groups. Our present research has resulted in the creation of precise prediction models that can effectively distinguish between individuals with DPN and those without DPN based on physiological characteristics. Additionally, we employed model-agnostic techniques to convert a black box model into a transparent model, allowing us to get insights into the underlying physiological characteristics of the two groups. We also used Qlattice symbolic regression method to develop transparent models exhibiting a non-linear relationship between DPN and the collective effect of Urea and Endocan, which are the two most important potential biomarkers.</div></div>","PeriodicalId":56010,"journal":{"name":"Egyptian Informatics Journal","volume":"31 ","pages":"Article 100777"},"PeriodicalIF":4.3,"publicationDate":"2025-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144922662","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Optimizing deep learning for webshell detection based on flexible dataset reduction","authors":"Saci Medileh ,&nbsp;Mohammad Hammoudeh ,&nbsp;Ahcene Bounceur ,&nbsp;Ferik Brahim ,&nbsp;Abdelkader Laouid ,&nbsp;Mostefa Kara ,&nbsp;Ammar Muthanna","doi":"10.1016/j.eij.2025.100770","DOIUrl":"10.1016/j.eij.2025.100770","url":null,"abstract":"<div><div>Webshells, malicious scripts, or code snippets have seen a dramatic rise in incidents, posing significant threats to organizations across various sectors. Traditional security measures often fail to detect these threats, necessitating the use of advanced detection mechanisms. This article proposes a deep learning-based technique for webshell detection, which addresses the challenges of high computational costs and sensitivity to input length variations. The proposed method uses a flexible dataset reduction approach in conjunction with two feature extraction techniques, TF-IDF and Word2Vec, to mitigate computational complexity and standardize model input. To address input variability and high-dimensionality, we introduce two dataset reduction strategies: Flat-based and Depth-based reduction, both of which rely on a standard deviation-based representation to preserve essential statistical characteristics while reducing dataset size. This combination enhances the performance and scalability of deep learning models, making them more feasible for practical applications in webshell detection. The study systematically reviews existing techniques, highlights limitations, and presents an innovative solution to improve detection accuracy and efficiency. Experimental results demonstrate that our approach achieves high accuracy (up to 98.50% using CNN) while significantly reducing training time. The findings validate that flexible dataset reduction combined with dual feature extraction offers a scalable and effective solution for real-time webshell detection.</div></div>","PeriodicalId":56010,"journal":{"name":"Egyptian Informatics Journal","volume":"31 ","pages":"Article 100770"},"PeriodicalIF":4.3,"publicationDate":"2025-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144919780","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Reliable generative interpretable framework for efficient predictive analysis of air quality index","authors":"M.K. Nallakaruppan ,&nbsp;C. Sai Varun ,&nbsp;Rajesh Kumar Dhanaraj ,&nbsp;Shashi Kant Tiwari ,&nbsp;V. Malathi ,&nbsp;Dragan Pamucar ,&nbsp;Dursun Delen","doi":"10.1016/j.eij.2025.100773","DOIUrl":"10.1016/j.eij.2025.100773","url":null,"abstract":"<div><div>Air quality management is one of the most important sustainability goals in the era of Industry 5.0. The magnitude of air pollution and impact of drastic pollutants increase day by day despite the significant efforts of the environmental enthusiasts and researchers. The role of Artificial Intelligence (AI) in determining the Air Quality Index (AQI) is significant with reasonable accuracy of classification achieved. The proposed model is a multi-class problem, that classifies the AQI into six different classes. Various ML models such as Decision Tree (DT), Random Forest (RF), Adaptive Boosting (AdaBoost), Gradient Boosting(GB), Logistic Regression (LR). The RF provided reliable performance metrics for AQI category prediction, achieving an accuracy and Precision of 0.99. This model is selected for the implementation of Explainable AI (XAI) models such as Local Interpretable Model Agonistic Explainer (LIME) for explanation using the local surrogacy plots and SHapley Additive exPlanations (SHAP) explainer for the global surrogacy plots. The Generative Adversarial Network (GAN) can generate synthetic data, which addresses critical issues such as missing data, class imbalance, noise, and redundant data. The performance the GAN shows optimized performance in classification of the AQI data with accuracy closer to 100 %. This is mainly due to the synthetic data generated by the GAN which enhances the performance of the classification. The proposed work integrates the efforts of the GAN-AI-XAI that enhances the performance, reliability, trustworthiness and robustness of the AQI classification model.</div></div>","PeriodicalId":56010,"journal":{"name":"Egyptian Informatics Journal","volume":"31 ","pages":"Article 100773"},"PeriodicalIF":4.3,"publicationDate":"2025-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144932240","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Traffic sign detection and recognition in Jordan based on machine learning and deep learning","authors":"Motasem S. Obeidat ,&nbsp;Ahmad H. Alomari ,&nbsp;Ameera S. Jaradat ,&nbsp;Malek M. Barhoush","doi":"10.1016/j.eij.2025.100761","DOIUrl":"10.1016/j.eij.2025.100761","url":null,"abstract":"<div><div>Traffic signs provide essential information to drivers, pedestrians, and cyclists on roads, highways, and other public areas, contributing significantly to road safety and order. This research investigates the effectiveness of a novel system for detecting and recognizing traffic signs in Jordan using machine learning and deep learning techniques. Specifically, we propose a methodology that integrates ResNet-50 for feature extraction with Support Vector Machine (SVM) for classification. This system leverages ResNet-50′s ability to extract intricate image features and SVM’s precision in classification tasks, achieving an impressive 83.05 % F1 score in recognizing various Jordanian traffic signs. The proposed approach provides a high-performing solution tailored to Jordanian road conditions, demonstrating that this combination of deep and machine learning techniques is effective for traffic sign recognition in real-world scenarios.</div></div>","PeriodicalId":56010,"journal":{"name":"Egyptian Informatics Journal","volume":"31 ","pages":"Article 100761"},"PeriodicalIF":4.3,"publicationDate":"2025-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144988581","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"IFC: Editorial","authors":"","doi":"10.1016/S1110-8665(25)00184-7","DOIUrl":"10.1016/S1110-8665(25)00184-7","url":null,"abstract":"","PeriodicalId":56010,"journal":{"name":"Egyptian Informatics Journal","volume":"31 ","pages":"Article 100791"},"PeriodicalIF":4.3,"publicationDate":"2025-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145121124","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Machine learning classification approaches for prediction of effective diabetes drugs","authors":"Ibrahim Abdelbaky ,&nbsp;Mariam Ahmed ,&nbsp;Mohamed Taha","doi":"10.1016/j.eij.2025.100786","DOIUrl":"10.1016/j.eij.2025.100786","url":null,"abstract":"<div><div>Diabetes, a complex and widespread metabolic disease, presents unique challenges for individuals and healthcare systems alike. This paper describes a model for personalized diabetes treatment by employing various classification approaches to assist medical professionals in accurately prescribing medications to patients. The primary objective was to predict the most appropriate drug treatment for individual patients by applying multi-label and multi- target classification techniques, we developed classification models that can improve the health of diabetic patients including predicting the risk of readmission for each patient by using two main approaches, the first approach is multi-label classification, this approach aimed to predict the most suitable drug treatment class for the patient. The second approach applied was multi-target classification, this approach will predict the most suitable drug treatment and the patient’s readmission. By considering multiple factors and characteristics specific to each patient, the model determined the suitable drug treatment based on their features and condition. To achieve a high-quality prediction of the suitable drug for diabetic patients, we employed feature engineering to enhance the efficiency and effectiveness of the machine learning algorithms used in the personalized treatment methodology. The experimental results indicate that the classification approaches are highly accurate when used to predict appropriate drug treatment for diabetes patients. The Naïve Bayes classifier reached an average accuracy of 98.72 %. Using cost-sensitive algorithms raised the average accuracy to 98 %.</div></div>","PeriodicalId":56010,"journal":{"name":"Egyptian Informatics Journal","volume":"31 ","pages":"Article 100786"},"PeriodicalIF":4.3,"publicationDate":"2025-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145048929","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Enhancing surveillance anomaly detection with keyframes and explainable inception model","authors":"Muhammad Salman ,&nbsp;Naveed Abbas ,&nbsp;Sayed Ijaz Ur Rahman ,&nbsp;Amjad Rehman ,&nbsp;Faten S. Alamri ,&nbsp;Alex Elyassih ,&nbsp;Tanzila Saba","doi":"10.1016/j.eij.2025.100769","DOIUrl":"10.1016/j.eij.2025.100769","url":null,"abstract":"<div><div>The proliferation of surveillance systems in the 21st century has become critical in mitigating rising crime rates and preserving life and property. However, conventional surveillance methods struggle with detecting abnormal human activities in complex environments characterized by varying lighting conditions and diverse appearances of individuals. To address these challenges, we propose advanced computer vision techniques, particularly supervised anomaly detection, to enhance intelligent surveillance systems. This research introduces SilentFrm to optimize anomaly detection by analyzing frame differences and setting thresholds based on calculated areas. SilentFrm streamlines the keyframe selection process, enhancing efficiency and reducing redundancy in anomaly detection tasks. Additionally, we present XAI-Inv3, an innovative architecture that integrates interpretability techniques such as Grad-CAM and guided backpropagation into the InceptionV3 model. These methodologies exhibit high accuracy and provide interpretable insights, thereby enhancing situational awareness and security. Rigorous experimentation showcases SilentFrm and XAI-Inv3′s superior detection accuracy and interpretability compared to state-of-the-art methods. By leveraging insights from weak supervision and semi-supervised learning, our approach aims to develop robust anomaly detection models capable of discerning anomalies with precision and efficiency in dynamic surveillance environments. Integrating XAI-Inv3 with the keyframes approach facilitates real-time anomaly detection and ensures scalability and feasibility across large-scale surveillance deployments. The proposed model achieved an average accuracy of 99 % on Hockey Fight, Violent Flow and Real-Life Violence Situation datasets for violence detection and 92 % for violence recognition using UCF-Crime and Shanghai Tech datasets. Proposed contributions underscore the importance of innovative methodologies in advancing the field of surveillance video analysis, ultimately enhancing public safety and crime prevention efforts.</div></div>","PeriodicalId":56010,"journal":{"name":"Egyptian Informatics Journal","volume":"31 ","pages":"Article 100769"},"PeriodicalIF":4.3,"publicationDate":"2025-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144922661","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Dam water levels prediction using advanced hybrid deep learning model based on Bayesian Optimization approach","authors":"Ahmet Enes Yegin ,&nbsp;Abdullah Ammar Karcioglu","doi":"10.1016/j.eij.2025.100760","DOIUrl":"10.1016/j.eij.2025.100760","url":null,"abstract":"<div><div>Effective dam water level prediction is of critical importance for the optimization of hydroelectric power generation, flood risk reduction and sustainable water resources management. In this study, a hybrid deep learning model is proposed for short-term water level prediction. In addition to deep learning models such as LSTM, BiLSTM, GRU and CNN, hybrid versions of these models (CNN-LSTM, CNN-BiLSTM, CNN-GRU) are also evaluated. The dataset used is based on daily hydrological data recorded between 2014 and 2023 of Deriner Dam, one of the strategically important dams of Turkey. The modeling process is supported by the Bayesian Optimization approach, which is one of the Neural Architecture Search (NAS) approaches, in order to minimize human intervention in hyperparameter selection. The NAS-optimized versions of each model are developed and compared separately. The highest accuracy was achieved with the proposed CNN-GRU Unified (CGU) hybrid model with a score of R² = 0.9941. The proposed CGU model combines spatial feature extraction and temporal dependencies modeling in the same structure, and better performance results are obtained with this model compared to state-of-the-art models and their hybrid versions. The high model accuracy and low error rate in the study show that the CGU architecture is a successful and reliable solution that can be integrated into real-time dam management systems. These findings have brought a new and scalable modeling approach to the literature, showing the usability of NAS-supported hybrid models in strategic water management applications.</div></div>","PeriodicalId":56010,"journal":{"name":"Egyptian Informatics Journal","volume":"31 ","pages":"Article 100760"},"PeriodicalIF":4.3,"publicationDate":"2025-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145048920","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"TransAdaptNet: a transformer-based adaptive learning model for accurate Monkeypox detection and classification","authors":"Dalia A. Magdi ,&nbsp;Ibrahim Obaya ,&nbsp;Fatma M. Talaat ,&nbsp;Warda M. Shaban","doi":"10.1016/j.eij.2025.100782","DOIUrl":"10.1016/j.eij.2025.100782","url":null,"abstract":"<div><div>The rise of Monkeypox (MPX) as a global health issue requires efficient and swift diagnostic techniques. In this paper, a Transformer-Based Adaptive Learning (TransAdaptNet) framework is proposed to enhance MPX detection and classification. The framework includes advanced AI methodologies, such as transformer-based models and adaptive learning systems, to significantly enhance the accuracy and efficiency of diagnosis. The proposed TransAdaptNet consists of two modules which are; data preprocessing, and patient classification. Through data preprocessing module, the used data set are preprocessed through several stages; remove/fill null values, remove outlier items, feature extraction and selection. A new feature selection methodology called Improved Whale Optimization Algorithm (IWOA)is introduced within the preprocessing pipeline. IWOA consists of two phases which are; Multi-Selection Phase (MSP) using two filter methods which are; fisher score and chi-square while Final Selection Phase (FSP) using Binary Whale Optimization Algorithm (BWOA) with union operations. Then, these features are fed into proposed patient classification module. TransAdaptNet achieved an exceptional accuracy of 98.7% utilizing a publicly accessible dataset comprising 500 monkeypox-positive and negative cases, surpassing conventional models like Random Forest and XGBoost. The framework has proven its ability to perform calculations quickly, avoiding the high costs associated with complex architectures such as transformers and attention mechanisms. TransAdaptNet produces clear outputs, enhancing the clarity of its predictions. The modular design ensures applicability in diverse healthcare settings, facilitating implementation. This method overcomes the limitations of traditional diagnostic tools, providing an effective and reliable means of identifying and mitigating MPX outbreaks at an early stage.</div></div>","PeriodicalId":56010,"journal":{"name":"Egyptian Informatics Journal","volume":"31 ","pages":"Article 100782"},"PeriodicalIF":4.3,"publicationDate":"2025-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145048927","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}