The compression sensing reconstruction for the 1-d signal can contribute to the communication of autonomous driving, intelligent robots, and fire exploration robots. To address the issue that fully connected layers in the LISTA method lack the ability to extract non-local features, this paper primarily designs a non-local fully connection layer and proposes a novel compressed sensing reconstruction method for audio signals. This paper designs a compression sensing reconstruction method for the 1-d signal. To reconstruct 1-d signal, the deep learning method LISTA is used. Then, the linear full connection layer in LISTA is improved by combining the output of three full connection layer to capture the non-local information. The computing regions of improved non-local full connection layer contain: 1) the full connection before; 2) the current full connection; and 3) the full connection after. Experimental results show the reconstruction results of LISTA and LISTA_nf are both close to the real signal. The MSE of LISTA_nf is reduced by 0.1 than the MSE of ISTA under the same experimental settings. The non-local full connection layer in the LISTA_nf consumes longer computing time. The LISTA_nf increase the computing time by 0.07s than the computing time of the ISTA. Experimental results show the effectiveness of the proposed method.

Under-five mortality remains a global health challenge with the rates of 43 deaths per every 1000 live births in Tanzania and 37 deaths per every 1000 live births globally. Although child mortality has significantly declined in the last twenty years, the current rates are far from reaching the anticipated Sustainable Development Goal of atmost 25 deaths per 1000 live births in 2030. This study intended to find the best performing classifier of under-five mortality status by comparing ten supervised machine learning algorithms. These machine learning algorithms are Decision Trees, Random Forest, Support Vector Machines, SMOTE-Based Boosted Random Forest, XGBoost, LightGBM, CatBoost, Logistic Regression, K-Nearest Neighbors and Stacked Ensemble Methods. The class imbalance of the dataset detected in the pre-processing stage was addressed using weighted categorical cross-entropy and SMOTE with a 5-folds cross validation and data splitting ratio of 80% for training set and 20% for testing set. With 20 experiments for each of the nine algorithms, the average results were reported to ensure that the findings were not by chance. Further, the stacking ensemble model was developed integrating six of the best performing algorithms using an inclusion criterion of AUC > 0.97. The findings revealed that ensemble algorithm consistently outperformed the other nine algorithms by achieving 100%, 100%, 99.97% and 99.24% for AUC, Accuracy, F1-Score and MCC respectively. This implies that stacking ensemble can uncover more insights than the individual algorithms in predicting under-five mortality status. This study recommends designing policies on under-five mortality that integrate insights from the stacking ensemble algorithm which shows the highest predictive performance.

This research paper focuses on detecting Cross-Site Scripting (XSS) attacks, a prevalent web security threat where attackers inject malicious scripts into web applications to steal sensitive user data, hijack sessions, and execute unauthorized actions. Traditional rule-based and signature-based detection methods often fail against sophisticated and obfuscated XSS payloads, necessitating more advanced solutions. To address this, a machine learning-based model is developed to enhance XSS detection accuracy while minimizing false positives. The proposed approach utilizes feature extraction techniques, including Term Frequency-Inverse Document Frequency (TF-IDF) and n-grams, to analyze JavaScript payloads, while Principal Component Analysis (PCA) is employed for feature selection, reducing dimensionality and improving computational efficiency. A Logistic Regression classifier is trained on an XSS payload dataset from Kaggle, with data split into 80% for training and 20% for testing to ensure a robust evaluation. Hyperparameter tuning is performed using GridSearchCV, optimizing the model’s predictive capabilities. Experimental results demonstrate a 99.70% accuracy, with 100% recall and 99.36% precision, highlighting the model’s effectiveness in detecting XSS attacks while minimizing false alarms. The high recall score ensures all malicious payloads are identified, while the strong precision rate enhances reliability for real-world deployment. These findings underscore the potential of machine learning in strengthening web security frameworks, offering a scalable and efficient alternative to conventional detection systems. Future research should focus on enhancing resilience against adversarial attacks by integrating deep learning models such as Bidirectional LSTMs (BiLSTMs) and Transformer-based architectures. Additionally, deploying the model in real-time web security solutions could provide proactive defense mechanisms, ensuring robust protection against evolving XSS threats.

The union combining the stitching of normal images and unsupervised semantic segmentation of the stitched image is an important region, which is crucial for autonomous driving, intelligent robots, and vehicle detection. This paper designs an union method combining the stitching of normal images and unsupervised semantic segmentation of the stitched image. The normal images are stitched by using the image stitching method designed by Ribeiro D.. The semantic segmentation method for the stitched image uses the method opened in the github. The stitched image contains image distortion. The distortion of the stitched image will make the feature extraction unreasonable. The distortion form of the stitched image is different from the distortion form of the panoramic image combined by line images. Therefore, the DCM proposed by Xing Hu is useless to extract features of the stitched image reasonable. This paper improves the DCM as the improved distortion convolution module (IDCM) by using the deformable convolution, the clamp module, the type transformation module, and the gather module. The IDCM is added before the unsupervised semantic segmentation method opened in the github to extract features reasonable. The IDCM-NUSSM method and the ISM-IDCM-NUSSM method are proposed. The experimental results show the better performance of the designed methods.

The advancement in the field of computer science, especially in machine learning (ML), represents a flourishing innovation that carries great importance in the domain of education. The beneficial impact of ML can also be observed in the realm of Qur’anic studies, particularly in Arabic text recognition and recitation analysis. This paper presents a comprehensive analysis of 34+ published scholarly articles devoted to Qur’anic studies. This work explores the convergence of machine learning methodologies and Qur’anic studies, examining the innovative applications and methodologies for Arabic text and voice classification. The fusion of ML algorithms makes the work easy and accurate to analyze, interpret, and extract valuable insights from the sacred text. Subsequently, we delve deeper into the emergent field of ML algorithms like k-NN, ANN, BLSTM, MFCC, SVM, NB and DL approaches have been adapted for Qur’anic texts classification, recitation and recitation analysis on accuracy, speed, class recognition, response rate and biasness benchmark. This work covers a diverse range of applications, including automated Qur’anic exegesis and analysis of usage of Ahkam Al-Tajweed. The main contribution of the work is to provide insight into how ML facilitates in Arabic and Kufic textual analysis, linguistic subtleties, and thematic structures of the Qur’anic text. Using the deep learning approaches, the reciters, recitation style and of the Quranic text has also explained in the work.

Amid the COVID-19 pandemic, extensive research has focused on deep learning methodologies for accurately diagnosing the virus from chest X-ray images. Various models, including Convolutional Neural Networks (CNNs) and pre-trained models, have achieved accuracies ranging from 85.20% to 99.66%. However, the proposed Fine-Tuned ResNet50 model consistently outperforms others with an impressive accuracy of 98.20%. By leveraging on transfer learning and careful architectural design the proposed model demonstrates superior performance compared to previous studies using DarkNet, ResNet50, and pre-trained models. Graphical comparisons highlight its competitive edge, emphasizing its effectiveness in COVID-19 classification tasks. The ResNet50 architecture, known for its deep residual layers and skip connections, facilitates robust feature extraction and classification, especially in medical imaging. Data pre-processing techniques, like noise reduction and contrast enhancement, ensure input data quality and reliability, enhancing the model's predictive abilities. Training results reveal the model's steady accuracy improvement and loss reduction over 20 epochs, aligning closely with validation metrics. Evaluation on a test set of COVID-19 chest X-ray images confirms exceptional accuracy (98.20%), precision (99.00%), recall (98.82%), and F1-score (98.91%), highlighting its proficiency in identifying COVID-19 cases while minimizing false positives and negatives. Comparative analyses against prior studies further validate its superior performance, establishing the Fine-Tuned ResNet50 model as a reliable tool for COVID-19 diagnosis. Future research should focus on exploring ensemble learning techniques, interpretability methods, and stakeholder collaboration to ensure safe AI deployment in clinical settings. Moreover, larger and diverse datasets are crucial for validating model performance and improving generalization, ultimately enhancing patient care and public health outcomes in the mitigating COVID-19 and future pandemics.

The development of technology today is growing so rapidly, almost all aspects of human life all use technology. The need for it-based teaching and learning concepts and mechanisms is increasingly advanced which then became known as E-learning brought the influence of the transformation of conventional education into digital forms, both in content (contens) and systems. E-learning is a learning model that utilizes information and communication technology facilities. With this information technology can act as the provision of information between students and students, learning resources and a means to streamline learning evaluation. SMP Negeri 2 Busalangga as one of the state schools whose teaching and learning activities process still use conventional means. To educate students or students in the face of national exams, teachers or schools always hold tutoring or try out, national pre-exams outside of school hours. A system is needed that can facilitate the teaching and learning process in schools between students and teachers, namely e-learning. The results obtained in the form of e-learning website media that can be used for teaching and learning activities where teachers and students can access the subject matter easily and quickly and can be accessed from anywhere and will be a plus because the system is based on information and technology.

The Internet of things, including Internet technology, including wired and wireless networks. Internet of Things and the Internet is the relationship between the parent and the child. In this paper, we aim to study the Investigation on the network packet loss’s long-range dependence and QOE and gain a good result and conclusion. In order to better establish no-reference video quality assessment model considering the network packet loss and further gain a better QoE evaluation, so we build NS2 + MyEvalvid simulation platform to study the scale characteristic of the network packet loss, scale characteristic of packet loss through the influence of packet loss rate to influence QoE. The experimental results show that, packet loss processes have long-range dependence, the number of superimposed source N, shape parameter, Hurst parameter, the output link speed have impacts on long-range dependence. We came to the conclusion that when superimposed source N is more, the shape parameter is smaller, Hurst parameter is bigger, the output link speed is smaller, packet loss’s long range dependence is larger, packet loss rate is high.

This paper carries on research surrounding the influences produced by dimensionality reduction on machine learning classification effect. Firstly, paper constructs the analysis architecture of data dimension reduction classification, combines the two different unsupervised dimension reduction methods, locally linear embedding (LLE) and principal component analysis (PCA) with the five machine learning classification methods: Gradient Boosting Decision Tree (GBDT), Random Forest, Support Vector Machine (SVM), K-Nearest Neighbor (KNN) and Logistic Regression. And then uses the handwritten digital identification dataset to analyze the classification performance of these five classification methods on different dimension datasets by different dimensionality reduction methods. The analysis shows that using the appropriate dimensionality reduction method for dimensionality reduction classification can effectively improve the classification accuracy; the dimensionality reduction classification effect of non-linear dimensionality reduction method is generally better than the linear dimensionality reduction method; different machine learning classification algorithms have significant differences in the sensitivity of dimensions.

This paper describes the translation algorithm of English negative sentences in rule-based English-Korean Machine Translation (EKMT). The proposed algorithm is based on the comparative study of the linguistic characteristics of English and Korean negative sentences. The earlier versions of machine translation system which is under development by our research team, failed to translate English negative sentences into accurate Korean equivalents. On the basis of the comparative linguistic research on negation in English and Korean, a new translation algorithm of the English negative sentences was established and evaluated.