Description
Making the world a safer place is not rocket science, it's data science.
In the field of cybersecurity, data scientists use computational methods to develop more effective ways to find security threats hidden in data flows of IT network communications. This talk, focused on data scientists, will provide a moderately technical representation of how machine learning and data science solve real-life problems, demonstrating how threat detection is enriched and improved through machine learning and data science.
I will address:
- The ups and downs of AI and ML in this application: What’s so complicated about cybersecurity anyway? * Machine Learning taxonomy: classification, clustering, anomaly detection
- When is an anomaly an anomaly, and when is it not an anomaly?
- How we built Machine Learning spaceships to identify one of the oldest and most persistent attack vectors (i.e. SQL Injections): Big, Bigger, and the Biggest versions.
- Trade-off with accuracy and complexity: Do we need to destroy a planet to get rid of SQL injections?