Peter Gyongyosi - Can You Trust UBA? -- Evaluating Machine Learning Algorithms in Practice
The promise of every security solution is to detect the next attack, but verifying that claim is almost impossible. Attacks are extremely rare and tend to change: the ability to catch attacks that happened in the past say little about the ability to find things that will happen in the future and those breached are unlikely to share information and data about how that happened. In this presentation I will show the different approaches and metrics we found to measure the efficiency of the unsupervised machine learning algorithms commonly used in UBA products.
Language: English • Duration: 17:35 • Resolution: 1280x720
Watch on YouTube Download the webcast Learn more about this conferenceVideo Links
Discover KuppingerCole
KuppingerCole PLUS
Get access to the whole body of KC PLUS research including Leadership Compass documents for only €800 a year
KuppingerCole Select
Register now for KuppingerCole Select and get your free 30-day access to a great selection of KuppingerCole research materials and to live trainings.