CBD: A Deep-Learning-Based Scheme for Encrypted Traffic Classification with a General Pre-Training Method.
Ontology highlight
ABSTRACT:
SUBMITTER: Hu X
PROVIDER: S-EPMC8705865 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:Traffic classification is essential in network-related areas such as network management, monitoring, and security. As the proportion of encrypted internet traffic rises, the accuracy of port-based and DPI-based traffic classification methods has declined. The methods based on machine learning and deep learning have effectively improved the accuracy of traffic classification, but they still suffer from inadequate extraction of traffic structure features and poor feature representativeness. This article proposes a model called Semi-supervision 2-Dimensional Convolution AutoEncoder (Semi-2DCAE). The model extracts the spatial structure features in the original network traffic by 2-dimensional convolution neural network (2D-CNN) and uses the autoencoder structure to downscale the data so that different traffic features are represented as spectral lines in different intervals of a one-dimensional standard coordinate system, which we call FlowSpectrum. In this article, the PRuLe activation function is added to the model to ensure the stability of the training process. We use the ISCX-VPN2016 dataset to test the classification effect of FlowSpectrum model. The experimental results show that the proposed model can characterize the encrypted traffic features in a one-dimensional coordinate system and classify Non-VPN encrypted traffic with an accuracy of up to 99.2%, which is about 7% better than the state-of-the-art solution, and VPN encrypted traffic with an accuracy of 98.3%, which is about 2% better than the state-of-the-art solution.
| S-EPMC10703001 | biostudies-literature
Project description:The virulence factor database (VFDB, http://www.mgc.ac.cn/VFs/) is dedicated to presenting a comprehensive knowledge base and a versatile analysis platform for bacterial virulence factors (VFs). Recent developments in sequencing technologies have led to increasing demands to analyze potential VFs within microbiome data that always consist of many different bacteria. Nevertheless, the current classification of VFs from various pathogens is based on different schemes, which create a chaotic situation and form a barrier for the easy application of the VFDB dataset for future panbacterial metagenomic analyses. Therefore, based on extensive literature mining, we recently proposed a general category of bacterial VFs in the database and reorganized the VFDB dataset accordingly. Thus, all known bacterial VFs from 32 genera of common bacterial pathogens collected in the VFDB are well grouped into 14 basal categories along with over 100 subcategories in a hierarchical architecture. The new coherent and well-defined VFDB dataset will be feasible and applicable for future panbacterial analysis in terms of virulence factors. In addition, we introduced a redesigned JavaScript-independent web interface for the VFDB website to make the database readily accessible to all users with various client settings worldwide.
| S-EPMC8728188 | biostudies-literature
Project description:We present a dataset that captures seven days of monitoring data from eight servers hosting more than 800 sites across a large campus network. The dataset contains data from network monitoring and host-based monitoring. The first set of data are packet traces collected by a probe situated on the network link in front of the web servers. The traces contain encrypted HTTP over TLS 1.2 communication between clients and web servers. The second set of data is an event log captured directly on the web servers. The events are generated by the Internet Information Services (IIS) logging and include both the IIS default features and custom features, such as client port and transferred data volume. Anonymization of all features in the dataset has been carefully carried out to prevent private information leakage while preserving the information value of the dataset. The dataset is suitable mainly for training machine learning techniques for anomaly detection and the identification of relationships between network traffic and events on web servers. We also add tools, settings, and a guide to convert the packet traces to IP flows that are often preferred for network traffic analysis.
| S-EPMC9065631 | biostudies-literature
Project description:Here we present miR-eCLIP analysis of AGO2 in HEK293 cells to address the small RNA repertoire and uncover their physiological targets. We developed an optimized bioinformatics approach of chimeric read identification to detect chimeras of high confidence, which were useed as an biologically validated input for miRBind, a deep learning method and web-server that can be used to accurately predict the potential of miRNA:target site binding.
2022-12-22 | GSE218466 | GEO
Project description:BackgroundDeep learning (DL) has shown promising results in molecular-based classification of glioma subtypes from MR images. DL requires a large number of training data for achieving good generalization performance. Since brain tumor datasets are usually small in size, combination of such datasets from different hospitals are needed. Data privacy issue from hospitals often poses a constraint on such a practice. Federated learning (FL) has gained much attention lately as it trains a central DL model without requiring data sharing from different hospitals.MethodWe propose a novel 3D FL scheme for glioma and its molecular subtype classification. In the scheme, a slice-based DL classifier, EtFedDyn, is exploited which is an extension of FedDyn, with the key differences on using focal loss cost function to tackle severe class imbalances in the datasets, and on multi-stream network to exploit MRIs in different modalities. By combining EtFedDyn with domain mapping as the pre-processing and 3D scan-based post-processing, the proposed scheme makes 3D brain scan-based classification on datasets from different dataset owners. To examine whether the FL scheme could replace the central learning (CL) one, we then compare the classification performance between the proposed FL and the corresponding CL schemes. Furthermore, detailed empirical-based analysis were also conducted to exam the effect of using domain mapping, 3D scan-based post-processing, different cost functions and different FL schemes.ResultsExperiments were done on two case studies: classification of glioma subtypes (IDH mutation and wild-type on TCGA and US datasets in case A) and glioma grades (high/low grade glioma HGG and LGG on MICCAI dataset in case B). The proposed FL scheme has obtained good performance on the test sets (85.46%, 75.56%) for IDH subtypes and (89.28%, 90.72%) for glioma LGG/HGG all averaged on five runs. Comparing with the corresponding CL scheme, the drop in test accuracy from the proposed FL scheme is small (-1.17%, -0.83%), indicating its good potential to replace the CL scheme. Furthermore, the empirically tests have shown that an increased classification test accuracy by applying: domain mapping (0.4%, 1.85%) in case A; focal loss function (1.66%, 3.25%) in case A and (1.19%, 1.85%) in case B; 3D post-processing (2.11%, 2.23%) in case A and (1.81%, 2.39%) in case B and EtFedDyn over FedAvg classifier (1.05%, 1.55%) in case A and (1.23%, 1.81%) in case B with fast convergence, which all contributed to the improvement of overall performance in the proposed FL scheme.ConclusionThe proposed FL scheme is shown to be effective in predicting glioma and its subtypes by using MR images from test sets, with great potential of replacing the conventional CL approaches for training deep networks. This could help hospitals to maintain their data privacy, while using a federated trained classifier with nearly similar performance as that from a centrally trained one. Further detailed experiments have shown that different parts in the proposed 3D FL scheme, such as domain mapping (make datasets more uniform) and post-processing (scan-based classification), are essential.
| S-EPMC10242007 | biostudies-literature
Project description:BACKGROUND:A national pre-registration pharmacist training recruitment scheme, which replaces local recruitment models, was introduced in England and Wales in 2017. The national recruitment system allows pharmacy students to apply for the 52?weeks training programmes (mandatory requirement for registration as a pharmacist), through a single application system prior to undertaking a nationally administered assessment. This study aimed to explore experiences of pharmacy students on the national recruitment scheme, particularly their views on the selection methodology, application process, and offer outcomes. METHODS:This mixed method study involved a) an online survey of all (approximate n =?2800) year 4 (final year of MPharm degree) pharmacy students in England and Wales and b) a qualitative focus group with four students. The study population was eligible to participate in the 2017/18 national recruitment scheme. Survey respondents were invited to participate in a focus group. Quantitative data were analysed using descriptive and inferential analysis. Qualitative data were analysed using the framework technique. Participation was voluntary. Ethical approval from University of Birmingham was obtained. RESULTS:A total of 307 completed surveys were returned (approximate response rate 11%). Respondents were generally satisfied with the application process and commended the fairness of the selection methodology and convenience in allowing them to apply to multiple training providers. Most survey respondents (n =?181, 72.9%) were either satisfied or highly satisfied with the training programme they were offered based on their assessment performances. Three themes and eight sub-themes obtained from the analysis of over 200 open comments data from the survey and transcript of a focus group with four participants. Results suggested the need to widen the timeframe available for applicants to shortlist their preferred employers, improve the method of programme listing in the application system, and consideration of prior achievements including academic performances and placement experiences to be included in the selection methodology. CONCLUSIONS:Experiences of pharmacy students on the national recruitment scheme suggest that respondents considered the selection methodology to be fair. Student engagement and satisfaction with the recruitment system can be maximised through improved listing of employers and widening the timescales for students to shortlist their preferred employers during application process. Inclusion of University achievements in the selection methodology will require consideration of evidence based approaches. Low response rate limits generalisation of findings.
| S-EPMC6894328 | biostudies-literature
Project description:MotivationClassification of samples using biomedical omics data is a widely used method in biomedical research. However, these datasets often possess challenging characteristics, including high dimensionality, limited sample sizes, and inherent biases across diverse sources. These factors limit the performance of traditional machine learning models, particularly when applied to independent datasets.ResultsTo address these challenges, we propose a novel classifier, Deep Centroid, which combines the stability of the nearest centroid classifier and the strong fitting ability of the deep cascade strategy. Deep Centroid is an ensemble learning method with a multi-layer cascade structure, consisting of feature scanning and cascade learning stages that can dynamically adjust the training scale. We apply Deep Centroid to three precision medicine applications-cancer early diagnosis, cancer prognosis, and drug sensitivity prediction-using cell-free DNA fragmentations, gene expression profiles, and DNA methylation data. Experimental results demonstrate that Deep Centroid outperforms six traditional machine learning models in all three applications, showcasing its potential in biological omics data classification. Furthermore, functional annotations reveal that the features scanned by the model exhibit biological significance, indicating its interpretability from a biological perspective. Our findings underscore the promising application of Deep Centroid in the classification of biomedical omics data, particularly in the field of precision medicine.Availability and implementationDeep Centroid is available at both github (github.com/xiexiexiekuan/DeepCentroid) and Figshare (https://figshare.com/articles/software/Deep_Centroid_A_General_Deep_Cascade_Classifier_for_Biomedical_Omics_Data_Classification/24993516).
| S-EPMC10868341 | biostudies-literature
Project description:The binding of microRNAs (miRNAs) to their target sites is a complex process, mediated by the Argonaute (Ago) family of proteins. The prediction of miRNA:target site binding is an important first step for any miRNA target prediction algorithm. To date, the potential for miRNA:target site binding is evaluated using either co-folding free energy measures or heuristic approaches, based on the identification of binding 'seeds', i.e., continuous stretches of binding corresponding to specific parts of the miRNA. The limitations of both these families of methods have produced generations of miRNA target prediction algorithms that are primarily focused on 'canonical' seed targets, even though unbiased experimental methods have shown that only approximately half of in vivo miRNA targets are 'canonical'. Herein, we present miRBind, a deep learning method and web server that can be used to accurately predict the potential of miRNA:target site binding. We trained our method using seed-agnostic experimental data and show that our method outperforms both seed-based approaches and co-fold free energy approaches. The full code for the development of miRBind and a freely accessible web server are freely available.
| S-EPMC9777820 | biostudies-literature
Project description:In the context of the proliferated evolution of network service types and the expeditious augmentation of network resource deployment, the requisition for copious labeled datasets to facilitate superior performance in traffic classification methods, particularly those hinging on deep learning, is imperative. Nonetheless, the procurement and annotation of such extensive datasets necessitate considerable temporal and human resource investments. In response to this predicament, this work introduces a methodology, termed MTEFU, leveraging a deep learning model-based multi-task learning algorithm, strategically designed to mitigate the reliance on substantial labeled training samples. Multiple classification tasks, encompassing duration, bandwidth size, and business traffic category, are incorporated, with a shared parameter strategy implemented amongst tasks to assure the transference of information across disparate tasks. Employing CNN, SAE, GRU, and LSTM as multi-task learning classification models, training validation and experimental testing were conducted on the QUIC dataset. A comparative analysis with single-task and ensemble learning methods reveals that, in the context of predicting network traffic types, the accuracy derived from the multi-task learning strategy, even with a mere 150 labeled samples, can emulate the 94.67% accuracy achieved through single-task learning with a fully labeled dataset of 6139 samples.
| S-EPMC10827795 | biostudies-literature
Project description:miRBind: a Deep Learning Method for miRNA Binding Classification
| PRJNA903902 | ENA