Traps and Pitfalls of Behavioural Analysis
Behavioural analysis of network traffic is progressively adopted by many security products and "machine learning" is the most used and mis-used buzzword in security today. The talk will start with elementary ideas behind machine learning methods. Then, we will present the differences between method classes and explain how they find attacks, why they cause false positives and how can the attackers intentionally or unintentionally avoid being detected by these methods. The talk will be on practitioner level, not assuming previous knowledge of the presented methods.
Petr Černohorský, Product Manager, Security Business Group at Cisco Systems
Cybersecurity expert from Cisco Advanced Threat Solutions BU, part of Cisco Security Business Group. He focuses on cloud-based cybersecurity, advanced threat landscape, behavior anomaly detection, advanced artificial intelligence and machine learning. He began his career at Cisco Systems through an acquisition of Cognitive Security in 2013. Petr regularly presents at cybersecurity conferences and publishes expert articles.