Cybersecurity is a critical domain as network attacks increasingly threaten systems worldwide. Detecting cyber attacks and analyzing anomalous behaviors in network traffic is essential for securing digital environments. This project leverages Recurrent Neural Networks (RNNs), technically Long Short-Term Memory (LSTM), to identify abnormal patterns in network traffic that signal potential cyberattacks, such as DDoS, port scanning, and brute-force attacks.
The project utilizes the CIC-IDS2017 Dataset, developed by the Canadian Institute for Cybersecurity. It is widely used for intrusion detection system evaluation and contains real-world attack simulations, including:
- DDoS attacks
- Brute force attacks
- SQL injection
- Port scanning
- Botnet activities
The dataset is suitable for RNN-based methods due to its sequential nature, allowing temporal dependencies in network traffic to be analyzed effectively.
For easier access, we used the Network Intrusion Dataset on Kaggle.
- To run this project, clone the repository and navigate to the project directory:
git clone https://github.com/nsswifter/CyberAttackDetection.git
cd CyberAttackDetection- Ensure you have
Python 3.11.1+and the required libraries installed. Use the command below to install dependencies:
pip install -r requirements.txt- Open the Jupyter Notebook:
jupyter notebook cyber_attack_detection.ipynb- Execute the notebook cells sequentially to preprocess data, train the model, and evaluate results.
The RNN model effectively detects cyber attacks with competitive accuracy, leveraging temporal patterns in the data. Detailed evaluation metrics and visualizations are included in the notebook.
This project is licensed under the MIT License.