Achieving accurate anomaly detection requires more than statistics. Simple assumptions like normal distribution do not work in the real world. Time series data – representing anything from customer acquisition, to application performance, to manufacturing KPIs – tend to have many different behaviors that need to be modeled accurately. These include seasonal patterns, non-stationary behaviors, and intricate correlations between signals, among others.
In this DSC webinar you will learn:
Fundamental machine learning techniques for anomaly detection
Requirements of an anomaly detection system in various use cases
Issues and pitfalls to watch out for when implementing anomaly detection
Common use cases and examples
Speaker: Ira Cohen, Chief Data Scientist — Anodot
Hosted by: Bill Vorhies, Editorial Director — Data Science Central
