Authors : Marin BOUTHEMY
This application written with Pygame aims to detect incoherent behavior during the user logging to a website. Based on the statistics related to user's taste and machine learning trained algorithm, it outputs a probability if the user trying to logging is a hacker (bot, fraud...) or not.
This application was developped for a hackathon challenge organized by Société Générale and HSBC in France in 2018.
It won the third prize corresponding to a 3000€ cash prize, you can find the results video here.
The library has some requirements :
- Python 3
- Numpy
- Pandas
- networkX
To install all this requirement you can use the following:
pip install requirements.txt
To run it for the sign up of the user (first connection):
python hacking_premiere_connexion.py
Then, you can activate it for the user sign in (second connection):
python hacking_seconde_connexion.py