Machine learning based Web Application firewall for detection attacks such as SQL injections, XSS and shell script injections.
Course project for TDA602 Contributors: Filip Granqvist & Oskar Holmberg
Requirements:
Follow along our development process in these notebooks:
To run notebooks (they can also be read from github):
Plots_technical_background.ipynb contains junk used to create images for the report
Demo-server contains a Node.js server with our best classifier (in the form of a .pickle) implemented available for live testing
See README.md in demo-server for instructions on how to set up the server
images contains images used for the report
data folder contains our malicious and non-malicious data. Also contains trained classifiers in form of .pickle files
tfidf_2grams_randomforest.p contains our single best classifier
trained_classifiers.p contains all our classifiers along with performance metrics. But this is not the final version, it was too big for github. Download the final version from here: https://1drv.ms/f/s!Aj1zBHCOJiQFgbQwDswBYtpzB1Pulg and replace the old one