Multi-Dimensional Network Anomaly Detection with Machine Learning
• Presentation
In this presentation, the authors introduce the state of the art in machine learning anomaly detection and give insight into techniques to limit the errors of statistical approaches.
Publisher
CounterFlow AI, Inc.
Topic or Tag
Abstract
In this presentation, the authors describe how recent multi-dimensional anomaly detection algorithms from machine learning can be used to combine traffic from multiple sources, while addressing the curse of dimensionality. Then, using an open-source platform of YAF, Apache Spark, and Apache Spot (incubating), they show how these algorithms can be used to provide effective focus for analysts and improve network outcomes.