Cyber Security: Power and Technology by Martti Lehto & Pekka Neittaanmäki

Author:Martti Lehto & Pekka Neittaanmäki
Language: eng
Format: epub
Publisher: Springer International Publishing, Cham

1.3 Contributions

We survey the existing machine learning based anomaly detection methods for DGA detection. We evaluate only techniques that are based on alphanumeric features of the domain names, without having to analyze any additional data, such as the IP address, the response packet to the DNS request, etc. This is more efficient and easy to implement than techniques that require this extra data.

Second, we introduce unsupervised anomaly detection methods for the first time in this field (to the best of our knowledge), which means that only a small portion of the data is required to be manually labeled in order to classify the domains as legitimate or malicious. Another important advantage of using unsupervised learning is the ability to detect DGAs that did not exist in the training data, as described in Sect. ‎5.4.1. We aim to prove the applicability of our unsupervised techniques by showing that the unsupervised K-Nearest Neighbors (KNN) method outperforms the previously used supervised ones.

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.