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:
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.