Repository of B.Tech Project on Botnet Detection using Network Traffic Behaviour Analysis and Machine Learning
Here we present Behavioral flow based Botnet detection approach using modern Machine Learning techniques such as Latest Classifiers and their combinations using Ensembling Techniques.
We also present a custom coded Flow Generator for Flow Identification and feature generation to characterize the network traffic.
Dataset used for analysis: Botnet Dataset by ISCX UNB, Canada
The description of the files and folders are:
- Botnet Docs contains some relevant Documents on Botnets and previous work
- Custom Flow Generator consists of a python implementation to extract the Bidirectional Traffic Flows and generate Flow Based Features to be used later for Machine Learning.
- ISCXFlowMeterMaster contains Flow generator given by ISCX. However the flow generator used in this project was custom written inline and also abstracted out for Traffic Analysis.
- Deep Learning Folder contains code for deep learning analysis of dataset and the results
- Rest are some code files done for analysis in no particular order