30min: "Stratosphere Project: Free Software Machine Learning to protect NGOs"
"NGOs and the civil society is at risk. They are continually being attacked by powerful actors because they are a highly political target. They don't have the resources to protect themselves with expensive solutions and they don't trust most companies. This is why the Czech Technical University started the Stratosphere IPS Project. It is an effort to develop a free software Network Intrusion Prevention System based on advanced Machine Learning algorithms. The goal of the project is to freely protect NGOs and the civil society at large by giving them this advanced defensive tool. The Stratosphere IPS Project provides an off-line free software tool and also a free cloud-based service where NGOs can send they network flows for real time detection. Stratosphere works by modeling the behavioral patterns of known malicious connections from malware and detecting new similar activities by using Markov Chains models. Stratosphere is free software because it is the best way to open our results and because NGOs should trust our work. This talk will show the Stratosphere project, the open model of sharing and how to detect real malware using machine learning in the network.
Our topic is a very important one: how the security tools should be free software if we are going to trust them. Most current solutions for using machine learning tools are proprietary because the algorithms are an asset of the company. In this business model, NGOs are left alone without any protection. The CTU University proposed to open its state-of-the-art research in order to protect NGOs with a trustable and advanced free software solution. We believe that NGOs need to work with organizations that are open about their work, about their mistakes and that are willing to show what they are doing. Currently Stratosphere is working with four organizations that are sending their data to us and we are doing our best to protect them continually.
Start time: 15:15
Click here to let us know how you liked this event.