UGR’16数据集
收藏资源简介:
UGR’16数据集是一个用于网络异常检测的数据集,由帕多瓦大学信息工程系提供。该数据集包括正常背景网络流量和异常流量,异常流量是通过将背景流量与使用高级黑客工具生成的攻击流量相结合而获得的。数据集分为校准集和测试集,校准集用于训练正常性模型,测试集包括干净流量和异常流量。异常流量分为拒绝服务攻击(DoS)和扫描攻击两种类型。数据集采用基于图像的表示方法,将网络流量信息映射到2D矩阵中,以简化处理流程并提高检测效率。
The UGR’16 dataset is a dataset for network anomaly detection, provided by the Department of Information Engineering of the University of Padua. This dataset contains both normal background network traffic and anomalous traffic. The anomalous traffic is obtained by combining the background traffic with attack traffic generated using advanced hacking tools. The dataset is divided into a calibration set and a test set, where the calibration set is used to train the normality model, and the test set includes both clean traffic and anomalous traffic. Anomalous traffic is categorized into two types: Denial of Service (DoS) attacks and scanning attacks. The dataset adopts an image-based representation method, which maps network traffic information into a 2D matrix to simplify the processing workflow and improve detection efficiency.
数据集概述
基本信息
- 标题: Unsupervised Network Anomaly Detection with Autoencoders and Traffic Images (EUSIPCO 2025)
- 作者:
- Michael Neri*(坦佩雷大学信息技术与通信科学学院,芬兰坦佩雷)
- Sara Baldoni°(帕多瓦大学信息工程系,意大利帕多瓦)
相关论文
- 论文标题: Unsupervised Network Anomaly Detection with Autoencoders and Traffic Images
- 会议: European Signal Processing Conference (EUSIPCO) 2025
- 引用格式: bibtex @INPROCEEDINGS{Neri_EUSIPCO_AD_2025, author={Neri, M. and Baldoni, S.}, booktitle={European Signal Processing Conference (EUSIPCO)}, title={{Unsupervised Network Anomaly Detection with Autoencoders and Traffic Images}}, year={2025}, volume={}, number={}, pages={}, doi= {}}
其他信息
- 用途: 该数据集用于基于自编码器和流量图像的无人监督网络异常检测研究。




