Inspiration
Our inspiration was being given an arduino uno Q, and with its network capabilities and compute power, we had INSPIRATION
What it does
Our project is a demonstration of how machine learning can be used to intercept and deflect malicious web traffic of different varieties.
How we built it
The project is written in python and utilizes pytorch to train on sample malicious and benign web traffic and later decide whether or not to allow the traffic through.
The setup is comprised of 2 servers, an attacker and defender. The attacker server has a preset list of malicious and benign web requests it can send at the defender at random. The defender utilized a multilayer perceptron network to decide whether or not to let the traffic through, and grabbed the traffic through scapy.
Challenges we ran into
Some notable challenges were:
- Wrestling the complicated properties of packets scapy presented to us
- Working with multiple computers to simulate the network attacks ## Accomplishments that we're proud of
- A very confident model and our ease of integration with other tools such as Tiger Data
Built With
- flask
- postgresql
- python
- pytorch
- scapy
- tigerdata
