The project aims to apply one machine learning (ML) technique to solve an issue in any field.
Distributed Denial of service (DDoS), attacks users from accessing the network and makes services unavailable or only partially available. The attacker floods the targeted machines or resources with excessive requests. The victim's incoming traffic originates from many different sources. Attackers use many (remotely) controlled computers to attack the victim.
In this project, we use the Naive Bayes technique to detect which if the request ddos or benign.
The Dataset is extracted from different IDS datasets that were produced in different years and different experimental DDoS traffic generation tools, it has more than 12 milion records (ddos and benign) and 85 features.
Required packages:
ipaddress
pip install ipaddress
seaborn
pip install seaborn
pandas
pip install pandas
matplotlib
pip install matplotlib
sklearn
pip install -U scikit-learn
After downloading "ProjectCode.zip" and packages, you have two options of environments to run:
In this option, you must have a large space to import and build a model according to a large dataset.
We use this option by using azure services in Machine Learning, is a cloud-based solution for ML workload and provides an end-to-end machine learning platform to enable users to build and deploy models faster on Azure. First, create a notebook for the project. Second, create a compute instance to execute the written code. The instance we chose had the following specifications Standard_E4a_v4 (4 cores, 32 GB RAM, 100 GB disk)
- Shoug Ali Alsuhaibani [email protected]
- Sarah Khalid Alaradi [email protected]