NetBench
收藏资源简介:
NetBench是一个大规模且全面的网络流量基准数据集,由威廉与玛丽学院计算机科学系创建,旨在评估机器学习模型,特别是基础模型在网络流量分类和生成任务中的表现。该数据集整合了7个公开数据集,涵盖了20个任务,包括15个分类任务和5个生成任务。数据集的创建过程涉及从原始网络流量数据中提取流和包,进行隐私保护处理,并使用统一的十六进制编码方法标准化数据格式。NetBench的应用领域广泛,主要用于网络流量的公平评估和基础模型的发展,以解决网络性能、安全和可靠性问题。
NetBench is a large-scale and comprehensive network traffic benchmark dataset, created by the Department of Computer Science at the College of William & Mary, aiming to evaluate machine learning models, especially foundation models, in network traffic classification and generation tasks. This dataset integrates seven public datasets, covering 20 tasks including 15 classification tasks and 5 generation tasks. The dataset creation process involves extracting flows and packets from raw network traffic data, conducting privacy-preserving processing, and standardizing the data format with a unified hexadecimal encoding method. NetBench has a wide range of application scenarios, and is primarily used for fair evaluation of network traffic and the development of foundation models, to address issues concerning network performance, security and reliability.

- 1NetBench: A Large-Scale and Comprehensive Network Traffic Benchmark Dataset for Foundation Models威廉与玛丽学院计算机科学系 · 2024年



