Evaluation of misbehavior detection mechanisms in Maat using VeReMi

This is the evaluation data for this paper: https://arxiv.org/abs/1804.06701 The analysis scripts for this data, as well as a brief explanation on what's what can be found here:<br> https://github.com/VeReMi-dataset/VeReMi-popper/tree/master/analysis This data was directly derived from the VeReMI dataset. Because the derivation scripts run on a cluster and are quite closely tied to that cluster, we decided to also publish the raw data. For processing, I split the data into 3 groups (low, medium, and high) of 75 tgz files that correspond to different densities (see paper for details). Basically this corresponds to A_R0_* to A_R14_*, A_R15_* to A_R29_* and A_R30_ to A_R44_* (this can also be clearly seen by the file size). If you store these in ./low, ./med, ./high respectively, you'll end up with the same graphs as published here: https://github.com/vs-uulm/securecomm2018-misbehavior-evaluation For the paper, we used a sub-set of these graphs.本数据集为论文https://arxiv.org/abs/1804.06701配套的评测数据。 本数据集的分析脚本及各文件的简要说明可在此处获取：https://github.com/VeReMi-dataset/VeReMi-popper/tree/master/analysis。 本数据集直接衍生自VeReMI数据集（VeReMI dataset）。由于衍生脚本需在集群环境运行且与该集群绑定度较高，我们决定同时发布原始数据。在处理环节，我们将数据划分为3组（低、中、高），每组包含75个tgz文件，分别对应不同的密度参数（详细说明参见论文）。具体而言，三组分别对应A_R0_*至A_R14_*、A_R15_*至A_R29_*以及A_R30_至A_R44_*（文件大小也可直观体现这一分组逻辑）。若将三组文件分别存放至./low、./med、./high目录下，即可复现此处发布的同款图表：https://github.com/vs-uulm/securecomm2018-misbehavior-evaluation。本论文仅使用了上述图表的一个子集。