This IDS project aims to detect and predict network intrusions using machine learning. The IDS is used to distinguish between 'bad connections' (intrusions/attacks) and a 'good (normal) connections' after applying some feature extraction on KDD Cup 1999 dataset by DARPA.

• Five different machine learning algorithms are used.
• The performance of all the algorithms is examined based on accuracy and computational times.
• KDD Cup 1999 dataset has been used for training and testing of the model.
• Derived results show that Random Forest outperforms the best on measures like Accuracy and Computational Time.

• Python 3.9 and Jupyter Notebook have been used.

1. Clone the repo

   git clone https://github.com/IUC4801/TDS.git

2. After cloning the repository, go to the IDS folder and open Jupyter Notebook.

The IDS was evaluated using KDD Cup 1999 dataset by DARPA dataset. The whole dataset can be downloaded from- http://kdd.ics.uci.edu/databases/kddcup99/kddcup99.html The model achieved an accuracy of 99.997% for Random Forest algorithm.

• OS: Windows 10 64 bit
• RAM: 8 GB
• Processor: 11th Gen Intel(R) Core(TM) i5