ICS-Flow
收藏资源简介:
ICS-Flow数据集是由瑞典梅拉达伦大学创新、设计与工程学院的研究人员创建,旨在为工业控制系统(ICS)的机器学习入侵检测技术提供评估基准。该数据集包含网络数据和过程状态变量日志,用于监督和非监督的机器学习入侵检测系统评估。数据集中的网络数据包括从模拟ICS组件和仿真网络捕获的正常和异常网络数据包和流量。异常是通过各种攻击技术注入系统的,这些技术常被黑客用于修改网络流量和破坏ICS。此外,研究团队还开发了开源工具“ICSFlowGenerator”,用于从原始网络数据包生成网络流量参数。最终数据集包含超过2500万个原始网络数据包、网络流量记录和过程变量日志。该数据集可用于训练入侵检测机器学习模型,并已公开在Kaggle平台上。
The ICS-Flow dataset was developed by researchers from the School of Innovation, Design and Engineering, Mälardalen University, Sweden, to serve as an evaluation benchmark for machine learning-based intrusion detection technologies in industrial control systems (ICS). This dataset includes network data and process state variable logs, which are used for evaluating both supervised and unsupervised machine learning intrusion detection systems. The network data within the dataset comprises normal and abnormal network packets and traffic captured from simulated ICS components and emulated networks. Abnormal instances are injected into the system via a variety of attack techniques commonly utilized by hackers to modify network traffic and compromise ICS. Furthermore, the research team developed an open-source tool named "ICSFlowGenerator" for generating network traffic parameters from raw network packets. The final dataset contains over 25 million raw network packets, network traffic records and process variable logs. This dataset can be employed to train machine learning models for intrusion detection, and has been publicly released on the Kaggle platform.




