CMPE 255 Project

Project proposal: https://docs.google.com/document/u/1/d/1Eae2wwwXqEORiP8mPCPLyD8CwjsVQ8aXj1zd8F0BXD0/edit?usp=drive_web&ouid=109387012218868591245
RandomForestAlgorithm implementation is based on Charanpal Dhanjal work https://gist.github.com/charanpald/c216800e25480ee838e8

Group members: Yang Chen, Fulbert Jong, David Tang

Environment setup

1. Install docker-ce
2. Install docker-compose via pip install docker-compose
3. The dataset used is the KDD Cup 99 dataset http://kdd.ics.uci.edu/databases/kddcup99/kddcup99.html

To download dataset:
Download KDD Cup 99 dataset with:

wget http://kdd.ics.uci.edu/databases/kddcup99/kddcup.data.gz

or

wget http://kdd.ics.uci.edu/databases/kddcup99/kddcup.data_10_percent.gz

and uncompress it. The first one occupies about 740MBytes, the second one around 71MBytes. Should you use the smaller dataset, please adjust filename in code: raw_data_filename = data_dir + "kddcup.data" change by raw_data_filename = data_dir + "kddcup.data_10_percent"

Running the project

docker-compose up --build
python detectAttack.py

Errors

1. ERROR: Couldn't connect to Docker daemon - you might need to run docker-machine start default.
   $newgrp docker