/neurotech-nids

DNN-powered network traffic analysis

Primary LanguagePythonMIT LicenseMIT

NeuroTech

NeuroTech is a network traffic analysis utility powered by deep learning with TensorFlow. With accuracy ratios of more than 85%, it's able to adapt to any environments and network conditions.

NeuroTech runs on Python 3.5.2. It also requires argus to be set up and running as a daemon on port 561, as well as a functioning ra client.

Module overview

learn

Generates datasets from current real network traffic under normal conditions.

train

Trains the neural network from a dataset (generated by learn or the base dataset).

test

Tests neural network accuracy against a test dataset. Shows the accuracy ratio.

capture

Captures and analyses current real network traffic.