PeerJ Computer Science最新文献

{"title":"Optimized hybrid SVM-RF multi-biometric framework for enhanced authentication using fingerprint, iris, and face recognition.","authors":"Sonal, Ajit Singh, Chander Kant","doi":"10.7717/peerj-cs.2699","DOIUrl":"10.7717/peerj-cs.2699","url":null,"abstract":"<p><p>This article introduces a hybrid multi-biometric system incorporating fingerprint, face, and iris recognition to enhance individual authentication. The system addresses limitations of uni-modal approaches by combining multiple biometric modalities, exhibiting superior performance and heightened security in practical scenarios, making it more dependable and resilient for real-world applications. The integration of support vector machine (SVM) and random forest (RF) classifiers, along with optimization techniques like bacterial foraging optimization (BFO) and genetic algorithms (GA), improves efficiency and robustness. Additionally, integrating feature-level fusion and utilizing methods such as Gabor filters for feature extraction enhances overall performance of the model. The system demonstrates superior accuracy and reliability, making it suitable for real-world applications requiring secure and dependable identification solutions.</p>","PeriodicalId":54224,"journal":{"name":"PeerJ Computer Science","volume":"11 ","pages":"e2699"},"PeriodicalIF":3.5,"publicationDate":"2025-02-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11888942/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143588196","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 lightweight coal-gangue detection model based on parallel deep residual networks.","authors":"Shexiang Jiang, Xinrui Zhou","doi":"10.7717/peerj-cs.2677","DOIUrl":"10.7717/peerj-cs.2677","url":null,"abstract":"<p><p>To realize the accurate identification of coal-gangue in the process of underground coal transportation and the low-cost deployment of the model, a lightweight coal-gangue detection model based on the parallel depth residual network, called P-RNet, is proposed. For the problem of images of coal-gangue taken under complex conditions, the feature extraction module (FEM) is designed using decoupling training and inference methods. Furthermore, for the problem of the nearest neighbor interpolation upsampling method being prone to produce mosaic blocks and edge jagged edges, a lightweight upsampling operator is used to optimize the feature fusion module (FFM). Finally, to solve the problem, the stochastic gradient descent algorithm is prone to local suboptimal solutions and saddle point problems in the error function optimization process, numerous experiments are carried out on selecting the initial learning rate, and the Lookahead optimizer is used to optimize parameters during backpropagation. Experimental results show that the proposed model can effectively improve the recognition effect, with a corresponding low deployment cost.</p>","PeriodicalId":54224,"journal":{"name":"PeerJ Computer Science","volume":"11 ","pages":"e2677"},"PeriodicalIF":3.5,"publicationDate":"2025-02-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11888869/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143588186","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":"An improved termite life cycle optimizer algorithm for global function optimization.","authors":"Yanjiao Wang, Mengjiao Wei","doi":"10.7717/peerj-cs.2671","DOIUrl":"10.7717/peerj-cs.2671","url":null,"abstract":"<p><p>The termite life cycle optimizer algorithm (TLCO) is a new bionic meta-heuristic algorithm that emulates the natural behavior of termites in their natural habitat. This work presents an improved TLCO (ITLCO) to increase the speed and accuracy of convergence. A novel strategy for worker generation is established to enhance communication between individuals in the worker population and termite population. This strategy would prevent the original worker generation strategy from effectively balancing algorithm convergence and population diversity to reduce the risk of the algorithm in reaching a local optimum. A novel soldier generation strategy is proposed, which incorporates a step factor that adheres to the principles of evolution to further enhance the algorithm's convergence speed. Furthermore, a novel replacement update mechanism is executed when the new individual is of lower quality than the original individual. This mechanism ensures a balance between the convergence of the algorithm and the diversity of the population. The findings from CEC2013, CEC2019, and CEC2020 test sets indicate that ITLCO exhibits notable benefits regarding convergence speed, accuracy, and stability in comparison with the basic TLCO algorithm and the four most exceptional meta-heuristic algorithms thus far.</p>","PeriodicalId":54224,"journal":{"name":"PeerJ Computer Science","volume":"11 ","pages":"e2671"},"PeriodicalIF":3.5,"publicationDate":"2025-02-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11888856/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143588245","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":"Combination of machine learning and data envelopment analysis to measure the efficiency of the Tax Service Office.","authors":"Shofinurdin Soffan, Arif Bramantoro, Ahmad A Alzahrani","doi":"10.7717/peerj-cs.2672","DOIUrl":"10.7717/peerj-cs.2672","url":null,"abstract":"<p><p>The Tax Service Office, a division of the Directorate General of Taxes, is responsible for providing taxation services to the public and collecting taxes. Achieving tax targets efficiently while utilizing available resources is crucial. To assess the performance efficiency of decision-making units (DMUs), data envelopment analysis (DEA) is commonly employed. However, ensuring homogeneity among the DMUs is often necessary and requires the application of machine learning clustering techniques. In this study, we propose a three-stage approach: Clustering, DEA, and Regression, to measure the efficiency of all tax service office units. Real datasets from Indonesian tax service offices were used while maintaining strict confidentiality. Unlike previous studies that considered both input and output variables, we focus solely on clustering input variables, as it leads to more objective efficiency values when combining the results from each cluster. The results revealed three clusters with a silhouette score of 0.304 and Davies Bouldin Index of 1.119, demonstrating the effectiveness of fuzzy c-means clustering. Out of 352 DMUs, 225 or approximately 64% were identified as efficient using DEA calculations. We propose a regression algorithm to measure the efficiency of DMUs in new office planning, by determining the values of input and output variables. The optimization of multilayer perceptrons using genetic algorithms reduced the mean squared error by about 75.75%, from 0.0144 to 0.0035. Based on our findings, the overall performance of tax service offices in Indonesia has reached an efficiency level of 64%. These results show a significant improvement over the previous study, in which only about 18% of offices were considered efficient. The main contribution of this research is the development of a comprehensive framework for evaluating and predicting tax office efficiency, providing valuable insights for improving performance.</p>","PeriodicalId":54224,"journal":{"name":"PeerJ Computer Science","volume":"11 ","pages":"e2672"},"PeriodicalIF":3.5,"publicationDate":"2025-02-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11888853/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143588335","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 multi-scale CNN with atrous spatial pyramid pooling for enhanced chest-based disease detection.","authors":"Muhammad Abdullah Shah Bukhari, Faisal Bukhari, Muhammad Asif, Hanan Aljuaid, Waheed Iqbal","doi":"10.7717/peerj-cs.2686","DOIUrl":"10.7717/peerj-cs.2686","url":null,"abstract":"<p><p>We introduce a sophisticated deep-learning model designed for the early detection of COVID-19 and pneumonia. The model employs a convolutional neural network-integrated with atrous spatial pyramid pooling. The atrous spatial pyramid pooling mechanism enhances the convolutional neural network model's ability to capture fine and large-scale features, optimizing detection accuracy in chest X-ray images. This improvement, along with transfer learning, significantly enhances the overall performance. By utilizing data augmentation to address the scarcity of available X-ray images, our atrous spatial pyramid pooling-enhanced convolutional neural network achieved a validation accuracy of 98.66% for COVID-19 and 83.75% for pneumonia, which beats the validation results of the other state of the art approaches (the metrics used for evaluation were accuracy, precision, F1-score, recall, specificity, and area under the curve). The model's multi-branch architecture facilitates more accurate and adaptable disease prediction, thereby increasing diagnostic precision and robustness. This approach offers the potential for faster and more reliable diagnoses of chest-related conditions.</p>","PeriodicalId":54224,"journal":{"name":"PeerJ Computer Science","volume":"11 ","pages":"e2686"},"PeriodicalIF":3.5,"publicationDate":"2025-02-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11888937/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143588113","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":"MFI-Net: multi-level feature invertible network image concealment technique.","authors":"Dapeng Cheng, Minghui Zhu, Bo Yang, Xiaolian Gao, Wanting Jing, Yanyan Mao, Feng Zhao","doi":"10.7717/peerj-cs.2668","DOIUrl":"10.7717/peerj-cs.2668","url":null,"abstract":"<p><p>The utilization of deep learning and invertible networks for image hiding has been proven effective and secure. These methods can conceal large amounts of information while maintaining high image quality and security. However, existing methods often lack precision in selecting the hidden regions and primarily rely on residual structures. They also fail to fully exploit low-level features, such as edges and textures. These issues lead to reduced quality in model generation results, a heightened risk of network overfitting, and diminished generalization capability. In this article, we propose a novel image hiding method based on invertible networks, called MFI-Net. The method introduces a new upsampling convolution block (UCB) and combines it with a residual dense block that employs the parametric rectified linear unit (PReLU) activation function, effectively utilizing multi-level information (low-level and high-level features) of the image. Additionally, a novel frequency domain loss (FDL) is introduced, which constrains the secret information to be hidden in regions of the cover image that are more suitable for concealing the data. Extensive experiments on the DIV2K, COCO, and ImageNet datasets demonstrate that MFI-Net consistently outperforms state-of-the-art methods, achieving superior image quality metrics. Furthermore, we apply the proposed method to digital collection images, achieving significant success.</p>","PeriodicalId":54224,"journal":{"name":"PeerJ Computer Science","volume":"11 ","pages":"e2668"},"PeriodicalIF":3.5,"publicationDate":"2025-02-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11888862/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143588169","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":"Construction of a user-friendly software-defined networking management using a graph-based abstraction layer.","authors":"Yufeng Jia, Jiadong Ren, Xianshan Li, Haitao He, Pengwei Zhang, Rong Li","doi":"10.7717/peerj-cs.2674","DOIUrl":"10.7717/peerj-cs.2674","url":null,"abstract":"<p><p>The software-defined networking (SDN) paradigm relies on the decoupling of the control plane and data plane. Northbound interfaces enable the implementation of network services through logical centralised control. Suitable northbound interfaces and application-oriented abstractions are the core of the SDN ecosystem. This article presents an architecture to represent the network as a graph. The purpose of this architecture is to implement an abstraction of the SDN controller at the application plane. We abstract all network elements using a graph model, with the attributes of the elements as the attributes of the graph. This virtualized logical abstraction layer, which is not limited by the physical network, enables network administrators to schedule network resources directly in a global view. The feasibility of the presented graph abstraction was verified through experiments in topological display, dynamic route, access control, and data persistence. The performance of the shortest path in the graph-based abstraction layer and graph database proves the necessity of the graph abstraction layer. Empirical evidence demonstrates that the graph-based abstraction layer can facilitate network slicing, maintain a dependable depiction of the real network, streamline network administration and network application development, and provide a sophisticated abstraction that is easily understandable to network administrators.</p>","PeriodicalId":54224,"journal":{"name":"PeerJ Computer Science","volume":"11 ","pages":"e2674"},"PeriodicalIF":3.5,"publicationDate":"2025-02-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11888870/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143588336","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":"Lung image segmentation with improved U-Net, V-Net and Seg-Net techniques.","authors":"Fuat Turk, Mahmut Kılıçaslan","doi":"10.7717/peerj-cs.2700","DOIUrl":"10.7717/peerj-cs.2700","url":null,"abstract":"<p><p>Tuberculosis remains a significant health challenge worldwide, affecting a large population. Therefore, accurate diagnosis of this disease is a critical issue. With advancements in computer systems, imaging devices, and rapid progress in machine learning, tuberculosis diagnosis is being increasingly performed through image analysis. This study proposes three segmentation models based on U-Net, V-Net, and Seg-Net architectures to improve tuberculosis detection using the Shenzhen and Montgomery databases. These deep learning-based methods aim to enhance segmentation accuracy by employing advanced preprocessing techniques, attention mechanisms, and non-local blocks. Experimental results indicate that the proposed models outperform traditional approaches, particularly in terms of the Dice coefficient and accuracy values. The models have demonstrated robust performance on popular datasets. As a result, they contribute to more precise and reliable lung region segmentation, which is crucial for the accurate diagnosis of respiratory diseases like tuberculosis. In evaluations using various performance metrics, the proposed U-Net and V-Net models achieved Dice coefficient scores of 96.43% and 96.42%, respectively, proving their competitiveness and effectiveness in medical image analysis. These findings demonstrate that the Dice coefficient values of the proposed U-Net and V-Net models are more effective in tuberculosis segmentation than Seg-Net and other traditional methods.</p>","PeriodicalId":54224,"journal":{"name":"PeerJ Computer Science","volume":"11 ","pages":"e2700"},"PeriodicalIF":3.5,"publicationDate":"2025-02-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11888921/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143588156","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":"Origin-destination prediction from road average speed data using GraphResLSTM model.","authors":"Guangtong Hu, Jun Zhang","doi":"10.7717/peerj-cs.2709","DOIUrl":"10.7717/peerj-cs.2709","url":null,"abstract":"<p><p>With the increasing demand for traffic management and resource allocation in Intelligent Transportation Systems (ITS), accurate origin-destination (OD) prediction has become crucial. This article presents a novel integrated framework, effectively merging the distinctive capabilities of graph convolutional network (GCN), residual neural network (ResNet), and long short-term memory network (LSTM), hereby designated as GraphResLSTM. GraphResLSTM leverages road average speed data for OD prediction. Contrary to traditional reliance on traffic flow data, road average speed data provides richer informational dimensions, reflecting not only vehicle volume but also indirectly indicating congestion levels. We use a real-world road network to generate road average speed data and OD data through simulations in Simulation of Urban Mobility (SUMO), thereby avoiding the influence of external factors such as weather. To enhance training efficiency, we employ a method combining the entropy weight method with the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) for key road segment selection. Using this generated dataset, carefully designed comparative experiments are conducted to compare various different models and data types. The results clearly demonstrate that both the GraphResLSTM model and the road average speed data markedly outperform alternative models and data types in OD prediction.</p>","PeriodicalId":54224,"journal":{"name":"PeerJ Computer Science","volume":"11 ","pages":"e2709"},"PeriodicalIF":3.5,"publicationDate":"2025-02-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11888923/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143588200","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":"Leveraging a hybrid convolutional gated recursive diabetes prediction and severity grading model through a mobile app.","authors":"Alhuseen Omar Alsayed, Nor Azman Ismail, Layla Hasan, Muhammad Binsawad, Farhat Embarak","doi":"10.7717/peerj-cs.2642","DOIUrl":"10.7717/peerj-cs.2642","url":null,"abstract":"<p><p>Diabetes mellitus is a common illness associated with high morbidity and mortality rates. Early detection of diabetes is essential to prevent long-term health complications. The existing machine learning model struggles with accuracy and reliability issues, as well as data imbalance, hindering the creation of a dependable diabetes prediction model. The research addresses the issue using a novel deep learning mechanism called convolutional gated recurrent unit (CGRU), which could accurately detect diabetic disorder and their severity level. To overcome these obstacles, this study presents a brand-new deep learning technique, the CGRU, which enhances prediction accuracy by extracting temporal and spatial characteristics from the data. The proposed mechanism extracts both the spatial and temporal attributes from the input data to enable efficient classification. The proposed framework consists of three primary phases: data preparation, model training, and evaluation. Specifically, the proposed technique is applied to the BRFSS dataset for diabetes prediction. The collected data undergoes pre-processing steps, including missing data imputation, irrelevant feature removal, and normalization, to make it suitable for further processing. Furthermore, the pre-processed data is fed to the CGRU model, which is trained to identify intricate patterns indicating the stages of diabetes. To group the patients based on their characteristics and identity patterns, the research uses the clustering algorithm which helps them to classify the severity level. The efficacy of the proposed CGRU framework is demonstrated by validating the experimental findings against existing state-of-the-art approaches. When compared to existing approaches, such as Attention-based CNN and Ensemble ML model, the proposed model outperforms conventional machine learning techniques, demonstrating the efficacy of the CGRU architecture for diabetes prediction with a high accuracy rate o f 99.9%. Clustering algorithms are more beneficial as they help in identifying the subtle pattern in the dataset. When compared to other methods, it can lead to more accurate and reliable prediction. The study highlights how the cutting-edge CGRU model enhances the early detection and diagnosis of diabetes, which will eventually lead to improved healthcare outcomes. However, the study limits to work on diverse datasets, which is the only thing considered to be the drawback of this research.</p>","PeriodicalId":54224,"journal":{"name":"PeerJ Computer Science","volume":"11 ","pages":"e2642"},"PeriodicalIF":3.5,"publicationDate":"2025-02-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11888868/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143588204","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}