Feature selection for HTTP botnet detection

Ang, Hooi Fang (2016) Feature selection for HTTP botnet detection. Project Report. Melaka, Malaysia, Universiti Teknikal Malaysia Melaka. (Submitted)

Text (Full Text) Feature selection for HTTP botnet detection.pdf - Submitted Version Download (1MB)

Abstract

HTTP botnet will be periodically visiting the certain web servers to obtain updates or new commands. HTTP protocol will be used by the botmasters to hide the malicious activities among the normal web flows. This will let the botmasters can easily evade current detection methods such as firewalls. Feature selection technique has been chosen to use in this project to detect the HTTP botnet. The scope of this project is focused on select the feature related to the HTTP botnet. The problem statement of this project is to analyse the difficulties to detect the botnet using network traffic due to the high number of parameters. This project has three objectives. The first objective is to study and investigate the network parameters. The second objective is to select the feature which is related to HTTP botnet using several selected feature selection techniques. The third objective is to compare the selected feature selection techniques used by choosing the more accuracy technique in feature selection of HTTP botnet. Project methodology can be divided into six phases which are Literature Review, Methodology, Data Collection, Pre-processing, Feature Selection, and Analysis and Documentation. The output of this project is to determine that some specific features is relevant to the HTTP botnet and select the best feature selection technique for HTTP botnet detection.

Actions (login required)

View Item