I developed a machine learning based malware classification system

Live application: [MDML](https://mdml.herokuapp.com/) (deployed on a free instance of Heroku, the performance is not impressive but it's good as a PoC, should be way faster locally or on a paid cloud instance)

the program has two main modules:

* **Static:** Analyzes the suspicious file statically without running it (only works with PE file like .exe and .dll)
* **Dynamic:** Runs the suspicious file in an isolated environment (sandbox) and extracts relevant features

both static and dynamic models are using XGBoost …

classification machine machine learning malware malware classification system