Machine Learning for Network Security and Malware Detection
You will create new machine learning algorithms to detect real malware traffic and attacks in the network. From studying the basis of network traffic, seeing real malware attack, create your dataset, extract features, to creating your own algorithm and evaluate it for improvements. A fast-paced hands-on training on how to get started in machine learning for network security.
To create new Machine Learning algorithms with the new python frameworks its easier than ever. However, our models still need designing, evaluation, tuning and specially good datasets and labels. In this training we will share high-quality and real network datasets of normal users being infected with malware. The goal is to learn to understand the problem, identify features, create your own ML models and finally test it against all the other models in the room! A fast-paced workshop going from traffic understanding to working python ML models in 2 days. Learn why ML is so difficult in network security and so useful. Work to obtain the highest detection performance and improve your knowledge. We will use many tools, including Colab, Python, NetFlows, scikit-learn and Pandas. At the end, take your own algorithm home!
The goals are: * To understand the basic principles of simple machine learning methods for network traffic. * To analyze real malware traffic and create in 2 days a machine learning algorithm to detect it. * To understand why ML is hard on network traffic and how to improve it * To dominate the basic tools and methods to get started with your own machine learning models
Attendees should know:
Network knowledge and protocols
Optionally is preferred:
- Basic Pandas
A web browser
A google account