Anomaly Detection
The Anomaly Detection Agent allows you to read sets of data and pinpoint anomalies using Microsoft Machine Learning algorithms. There are two use cases for this Agent:
Change Point Detection predicts when a value may indicate a new trend. For instance, pinpointing a rising or falling trend in the average electricity consumption of a device.
Spike Detection flags whether the difference between values is anomalously large. For instance, raising an alert when the running temperature of a device suddenly increases beyond the fluctuations expected in normal use.
The Anomaly Detection Agent currently supports two customizable algorithms to implement these use cases: one based on Independent and Identically Distributed values (IID), and another based on Singular Spectrum Analysis (SSA).
Details for an example and its configuration can be found in the How to Use section.
For documentation on Microsoft's ML .NET, see here.
Pre-requisites
The minimum XMPro Data Stream version required for this Agent is 4.1.
The Stream Host that this Agent will run on must have a Windows operating system.
Download
Please contact XMPro if you're looking for an older version of this Agent.
Last updated