An empirical comparison of botnet detection methods

摘要: The results of botnet detection methods are usually presented without any comparison. Although it is generally accepted that more comparisons with third-party may help to improve the area, few papers could do it. Among factors prevent a comparison difficulties share dataset, lack good absence proper description and methodology. This paper compares output three different by executing them over new, real, labeled large dataset. dataset includes botnet, normal background traffic. our two (BClus CAMNEP) BotHunter were compared using methodology novel error metric designed for detections methods. We conclude comparing indeed helps better estimate how are, algorithms, build datasets