PeerJ Computer Science最新文献

{"title":"MSKT: multimodal data fusion for improved nursing management in hemorrhagic stroke.","authors":"Ting Zhou, Dandan Li, Jingfang Zuo, Aihua Gu, Li Zhao","doi":"10.7717/peerj-cs.2969","DOIUrl":"10.7717/peerj-cs.2969","url":null,"abstract":"<p><strong>Background: </strong>The study aims to address the challenges of nursing decision-making and the optimization of personalized nursing plans in the management of hemorrhagic stroke. Due to the rapid progression and high complexity of hemorrhagic stroke, traditional nursing methods struggle to cope with the challenges posed by its high incidence and high disability rate.</p><p><strong>Methods: </strong>To address this, we propose an innovative approach based on multimodal data fusion and a non-stationary Gaussian process model. Utilizing multidimensional data from the MIMIC-IV database (including patient medical history, nursing records, laboratory test results, <i>etc</i>.), we developed a hybrid predictive model with a multiscale kernel transformer non-stationary Gaussian process (MSKT-NSGP) architecture to handle non-stationary time-series data and capture the dynamic changes in a patient's condition.</p><p><strong>Results: </strong>The proposed MSKT-NSGP model outperformed traditional algorithms in prediction accuracy, computational efficiency, and uncertainty handling. For hematoma expansion prediction, it achieved 85.5% accuracy, an area under the curve (AUC) of 0.87, and reduced mean squared error (MSE) by 18% compared to the sparse variational Gaussian process (SVGP). With an inference speed of 55 milliseconds per sample, it supports real-time predictions. The model maintained a confidence interval coverage near 95% with narrower widths, indicating precise uncertainty estimation. These results highlight its potential to enhance nursing decision-making, optimize personalized plans, and improve patient outcomes.</p>","PeriodicalId":54224,"journal":{"name":"PeerJ Computer Science","volume":"11 ","pages":"e2969"},"PeriodicalIF":3.5,"publicationDate":"2025-06-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12193459/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144499288","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Design of an improved graph-based model integrating LSTM, LoRaWAN, and blockchain for smart agriculture.","authors":"Ravi Kumar Munaganuri, Narasimha Rao Yamarthi, Sai Chandana Bolem","doi":"10.7717/peerj-cs.2896","DOIUrl":"https://doi.org/10.7717/peerj-cs.2896","url":null,"abstract":"<p><p>This research is anchored on the burning need for irrigation optimization and crop water use efficiency improvement, which remains a challenge in smart agriculture processes. Traditional irrigation methods normally lead to inefficiency, resulting in wasted water and non-maximum crops. These traditional ways normally lack attributes of real-time adaptability and secure data management-things that are very key to modernizing agricultural practices. In this work, artificial intelligence (AI), Internet of Things (IoT), and blockchain techniques will be integrated to design a comprehensive system for monitoring and predicting soil moisture levels. In the proposed model, long short-term memory (LSTM) networks are considered for soil moisture level prediction, taking into consideration past data, weather, and crop type. LSTM networks are chosen here for their high performance in timestamp series prediction tasks with an mean average error (MAE) of 0.02 m³/m³ over a 7-day forecast horizon. For real-time monitoring, IoT sensors based on long range wide area network (LoRaWAN) technology are field-deployed for conducting long-range communications while consuming very limited energy to extend the sensor battery life over 5 years and bring down the data transmission latency below 5 s. It has an inbuilt permissioned blockchain framework-Hyperledger Fabric-which offers a secure and transparent system for data management and maintaining a record of soil moisture data, irrigation events, and metadata from sensors. This ensures the immutability and integrity of sets of data. Smart contracts automate irrigation upon reaching preconfigured soil moisture thresholds, and hence zero data integrity breaches occur with a transaction throughput of 1,000 transactions per second, taken into view with smart contract execution latency of less than 2 s. Moreover, it utilizes reinforcement learning with Deep Q-Learning to derive an optimized irrigation schedule. In this regard, it enables learning optimal irrigation policies and implements them to improve efficiency in the usage of water by 25% and increases crop yield by 15% compared to the traditional methods. Clearly from field trials, results indicate evident efficiency of the integrated system: a 20% water usage reduction and a 12% increase in crop yield within one growing season. This is rather an innovative take on irrigation practices, increasing a great deal of accuracy and sustainability for such and providing a really strong solution toward better agricultural productivity and resource management.</p>","PeriodicalId":54224,"journal":{"name":"PeerJ Computer Science","volume":"11 ","pages":"e2896"},"PeriodicalIF":3.5,"publicationDate":"2025-06-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12192981/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144499242","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Optimized customer churn prediction using tabular generative adversarial network (GAN)-based hybrid sampling method and cost-sensitive learning.","authors":"I Nyoman Mahayasa Adiputra, Paweena Wanchai, Pei-Chun Lin","doi":"10.7717/peerj-cs.2949","DOIUrl":"10.7717/peerj-cs.2949","url":null,"abstract":"<p><strong>Background: </strong>Imbalanced and overlapped data in customer churn prediction significantly impact classification results. Various sampling and hybrid sampling methods have demonstrated effectiveness in addressing these issues. However, these methods have not performed well with classical machine learning algorithms.</p><p><strong>Methods: </strong>To optimize the performance of classical machine learning on customer churn prediction tasks, this study introduces an extension framework called CostLearnGAN, a tabular generative adversarial network (GAN)-hybrid sampling method, and cost-sensitive Learning. Utilizing a cost-sensitive learning perspective, this research aims to enhance the performance of several classical machine learning algorithms in customer churn prediction tasks. Based on the experimental results classical machine learning algorithms exhibit shorter execution times, making them suitable for predicting churn in large customer bases.</p><p><strong>Results: </strong>This study conducted an experiment with six comparative sampling methods, six datasets, and three machine learning algorithms. The results show that CostLearnGAN achieved a satisfying result across all evaluation metrics with a 1.44 average mean rank score. Additionally, this study provided a robustness measurement for algorithms, demonstrating that CostLearnGAN outperforms other sampling methods in improving the performance of classical machine learning models with a 5.68 robustness value on average.</p>","PeriodicalId":54224,"journal":{"name":"PeerJ Computer Science","volume":"11 ","pages":"e2949"},"PeriodicalIF":3.5,"publicationDate":"2025-06-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12193428/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144499301","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Sporting a virtual future: exploring sports and virtual reality patents using deep learning-based analysis.","authors":"Jea Woog Lee, Sangmin Song, JungMin Yun, Doug Hyun Han, YoungBin Kim","doi":"10.7717/peerj-cs.2919","DOIUrl":"10.7717/peerj-cs.2919","url":null,"abstract":"<p><p>We investigate the convergence of sports and emerging technologies from the Fourth Industrial Revolution, with a focus on virtual reality (VR) applications. Using patent big data, we introduce SportsBERT, a bidirectional encoder representation from transformers (BERT)-based algorithm tailored for enhanced natural language processing in sports-related knowledge-based documents. Through topic modeling, we extract key themes and clusters from sports-related VR patents, providing insights into the knowledge structure and technological trends in VR applications for sports. Our analysis identifies key drivers of technological advancement, including spatial hardware, tactile human-computer interactions, aerobic exercise, rehabilitation, and swing sports. Additionally, we highlight challenges such as the high cost and usability limitations of current VR devices. This study presents the first deep learning-based topic modeling approach specialized for sports patents and offers a comprehensive roadmap for current developments and future trajectories in VR sports technologies.</p>","PeriodicalId":54224,"journal":{"name":"PeerJ Computer Science","volume":"11 ","pages":"e2919"},"PeriodicalIF":3.5,"publicationDate":"2025-06-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12192641/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144499364","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Topic adversarial neural network for cross-topic cyberbullying detection.","authors":"Shufeng Xiong, Wenzhuo Liu, Bingkun Wang, Yinchao Che, Lei Shi","doi":"10.7717/peerj-cs.2942","DOIUrl":"10.7717/peerj-cs.2942","url":null,"abstract":"<p><p>With the proliferation of social media, cyberbullying has emerged as a pervasive threat, causing significant psychological harm to individuals and undermining social cohesion. Its linguistic expressions vary widely across topics, complicating automatic detection efforts. Most existing methods struggle to generalize across diverse online contexts due to their reliance on topic-specific features. To address this issue, we propose the Topic Adversarial Neural Network (TANN), a novel end-to-end framework for topic-invariant cyberbullying detection. TANN integrates a multi-level feature extractor with a topic discriminator and a cyberbullying detector. It leverages adversarial training to disentangle topic-related information while retaining universal linguistic cues relevant to harmful content. We construct a multi-topic dataset from major Chinese social media platforms, such as Weibo and Tieba, to evaluate the generalization performance of TANN in real-world scenarios. Experimental results demonstrate that TANN outperforms existing methods in cross-topic detection tasks, significantly improving robustness and accuracy. This work advances cross-topic cyberbullying detection by introducing a scalable solution that mitigates topic interference and enables reliable performance across dynamic online environments.</p>","PeriodicalId":54224,"journal":{"name":"PeerJ Computer Science","volume":"11 ","pages":"e2942"},"PeriodicalIF":3.5,"publicationDate":"2025-06-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12193452/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144499376","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A methodological approach for inferring causal relationships from opinions and news-derived events with an application to climate change.","authors":"Juan Marten, Fernando Delbianco, Fernando Tohme, Ana G Maguitman","doi":"10.7717/peerj-cs.2964","DOIUrl":"10.7717/peerj-cs.2964","url":null,"abstract":"<p><p>Social media platforms like Twitter (now X) provide a global forum for discussing ideas. In this work, we propose a novel methodology for detecting causal relationships in online discourse. Our approach integrates multiple causal inference techniques to analyze how public sentiment and discourse evolve in response to key events and influential figures, using five causal detection methods: Direct-LiNGAM, PC, PCMCI, VAR, and stochastic causality. The datasets contain variables, such as different topics, sentiments, and real-world events, among which we seek to detect causal relationships at different frequencies. The proposed methodology is applied to climate change opinions and data, offering insights into the causal relationships among public sentiment, specific topics, and natural disasters. This approach provides a framework for analyzing various causal questions. In the specific case of climate change, we can hypothesize that a surge in discussions on a specific topic consistently precedes a change in overall sentiment, level of aggressiveness, or the proportion of users expressing certain stances. We can also conjecture that real-world events, like natural disasters and the rise to power of politicians leaning towards climate change denial, may have a noticeable impact on the discussion on social media. We illustrate how the proposed methodology can be applied to examine these questions by combining datasets on tweets and climate disasters.</p>","PeriodicalId":54224,"journal":{"name":"PeerJ Computer Science","volume":"11 ","pages":"e2964"},"PeriodicalIF":3.5,"publicationDate":"2025-06-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12193450/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144499143","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Novel conditional tabular generative adversarial network based image augmentation for railway track fault detection.","authors":"Ali Raza, Rukhshanda Sehar, Abdul Moiz, Ala Saleh Alluhaidan, Sahar A El-Rahman, Diaa Salama AbdElminaam","doi":"10.7717/peerj-cs.2898","DOIUrl":"10.7717/peerj-cs.2898","url":null,"abstract":"<p><p>Railway track fault recognition is a critical aspect of railway maintenance, aiming to identify and rectify defects such as cracks, misalignments, and wear on tracks to ensure safe and efficient train operations. Classical methods for fault detection, including manual inspections and simple sensor-based systems, face significant challenges, such as high labour costs, human error, and limited detection accuracy under varying environmental conditions. These methods are often time-consuming and unable to provide real-time monitoring, leading to potential safety risks and operational inefficiencies. To address these challenges, efficient artificial intelligence-based image classification is being explored to enhance railway track fault detection accuracy, efficiency, and reliability. This research aims to develop an advanced generative neural network for efficient railway track fault detection. We propose a novel conditional tabular generative adversarial network (CTGAN)-based image augmentation approach to producing realistic synthetic image data using railway track images. We developed five advanced neural network techniques for comparison with railway track image classification. The random forest approach surpasses state-of-the-art studies with a high accuracy score of 0.99 for railway track fault detection. Hyperparameter optimization is applied to achieve optimal performance, and the performance is evaluated using the k-fold cross-validation approach. The proposed research enhances operational efficiency, reduces maintenance costs, and significantly improves the safety and reliability of rail transportation.</p>","PeriodicalId":54224,"journal":{"name":"PeerJ Computer Science","volume":"11 ","pages":"e2898"},"PeriodicalIF":3.5,"publicationDate":"2025-06-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12192886/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144499297","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Automated guided vehicle (AGV) path optimization method based on improved rapidly-exploring random trees.","authors":"Zhigang Ren, Anjiang Cai, Feilong Xu","doi":"10.7717/peerj-cs.2915","DOIUrl":"10.7717/peerj-cs.2915","url":null,"abstract":"<p><p>In response to the issues of low computational efficiency, slow convergence speed, curvy paths, and the tendency to fall into local optima in rapidly-exploring random tree (RRT) algorithms for automated guided vehicle (AGV) path planning, this article proposes an improved RRT algorithm that combines adaptive step-size optimization with K-dimensional tree (KD-Tree) based fast nearest neighbor search. Firstly, an adaptive step-size optimization strategy is introduced to dynamically adjust the step size during node searches, improving both the planning quality and computational efficiency of the algorithm. Secondly, the KD-Tree nearest neighbor search method is employed to accelerate node searching and reduce the time cost of path planning. Finally, a cubic spline interpolation function is applied to smooth the optimal path, further enhancing the planning quality. Experimental results show that the improved RRT algorithm significantly outperforms traditional RRT, RRT*, and Informed-RRT* in terms of path length, planning time, and path smoothness. Specifically, the average path length is reduced by 164.33 m, and the average search time is shortened by 3.3 s, making it more suitable for AGV path planning in practical applications.</p>","PeriodicalId":54224,"journal":{"name":"PeerJ Computer Science","volume":"11 ","pages":"e2915"},"PeriodicalIF":3.5,"publicationDate":"2025-06-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12192648/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144499215","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A new approach of anomaly detection in shopping center surveillance videos for theft prevention based on RLCNN model.","authors":"Muhammad Sajid, Ali Haider Khan, Kaleem Razzaq Malik, Javed Ali Khan, Ayed Alwadain","doi":"10.7717/peerj-cs.2944","DOIUrl":"10.7717/peerj-cs.2944","url":null,"abstract":"<p><p>The amount of video data produced daily by today's surveillance systems is enormous, making analysis difficult for computer vision specialists. It is challenging to continuously search these massive video streams for unexpected accidents because they occur seldom and have little chance of being observed. Contrarily, deep learning-based anomaly detection decreases the need for human labor and has comparably trustworthy decision-making capabilities, hence promoting public safety. In this article, we introduce a system for efficient anomaly detection that can function in surveillance networks with a modest level of complexity. The proposed method starts by obtaining spatiotemporal features from a group of frames. The multi-layer extended short-term memory model can precisely identify continuing unusual activity in complicated video scenarios of a busy shopping mall once we transmit the in-depth features extracted. We conducted in-depth tests on numerous benchmark datasets for anomaly detection to confirm the proposed framework's functionality in challenging surveillance scenarios. Compared to state-of-the-art techniques, our datasets, UCF50, UCF101, UCFYouTube, and UCFCustomized, provided better training and increased accuracy. Our model was trained for more classes than usual, and when the proposed model, RLCNN, was tested for those classes, the results were encouraging. All of our datasets worked admirably. However, when we used the UCFCustomized and UCFYouTube datasets compared to other UCF datasets, we achieved greater accuracy of 96 and 97, respectively.</p>","PeriodicalId":54224,"journal":{"name":"PeerJ Computer Science","volume":"11 ","pages":"e2944"},"PeriodicalIF":3.5,"publicationDate":"2025-06-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12193423/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144499144","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Enhancing analogy-based software cost estimation using Grey Wolf Optimization algorithm.","authors":"Taghi Javdani Gandomani, Maedeh Dashti, Sadegh Ansaripour, Hazura Zulzalil","doi":"10.7717/peerj-cs.2794","DOIUrl":"10.7717/peerj-cs.2794","url":null,"abstract":"<p><p>Accurate software cost estimation (SCE) is a critical factor in the successful delivery of software projects, as highlighted by industry statistics indicating that only some of the projects comply with the predicted budget. Among the software estimation methods, analogy-based estimation (ABE) is one of the most popular ones. Although this method has been customized in recent years with the help of optimization algorithms to achieve better results, the use of more powerful optimization algorithms can be effective in achieving better results in software size estimation. This study presents an innovative approach to SCE that integrates the grey wolf optimization (GWO) algorithm to enhance the precision of ABE. The GWO algorithm, inspired by the hunting behavior and social hierarchy of grey wolves, is mathematically modeled and incorporated into the ABE approach. The key focus of this research is the optimization of the similarity function, a crucial component of the ABE, using both Euclidean and Manhattan distance measures. The article addresses the challenges in selecting an optimal similarity function and emphasizes the importance of proper feature weighting to improve estimation accuracy. The proposed GWO-based ABE method is rigorously evaluated on multiple software project datasets using cross-validation techniques. The performance of the GWO-based ABE is compared to other evolutionary algorithms based on widely accepted evaluation metrics. The results confirm that the integration of the GWO algorithm into ABE enhances estimation accuracy and model robustness. By optimizing feature weights in the similarity function, GWO-ABE effectively addresses key limitations of traditional analogy-based methods. The proposed approach demonstrates superior performance across multiple datasets, particularly under the Euclidean distance function, making it a reliable solution for software project cost estimation. Experimental evaluations show that GWO-ABE achieves notable improvements in key performance metrics, leading to reduced mean magnitude of relative error (MMRE), median magnitude of relative error (MdMRE), and higher percentage of prediction (PRED) compared to other ABE-customized methods. These findings highlight the role of metaheuristic optimization in improving software estimation techniques, contributing to more precise and efficient project planning and management.</p>","PeriodicalId":54224,"journal":{"name":"PeerJ Computer Science","volume":"11 ","pages":"e2794"},"PeriodicalIF":3.5,"publicationDate":"2025-06-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12190706/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144499268","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}