Validating Machine Learning and Security Analytics
It is well understood that data is the key for proper development and maturity of any analytics or machine learning tool. Meaningful domain-specific data, events, or logs help to establish the baseline behavior of a domain, and eventually helps in understanding the tool’s ability to detect anomalies that diverge from the baseline behavior. A universal truth of analytics and machine learning is that the larger the volume of baseline data a model receives, the better it gets.
This document describes the challenges of validating using a diverse set of traffic patterns that replicate behaviors of different types of domains and vertical markets, and how Ixia's BreakingPoint will help you overcome them.