Streaming Detection

Streaming anomaly detection identifies real-time anomalies in continuous data, enabling immediate response to issues in dynamic systems.

What is it for?

Streaming anomaly detection is used to identify unusual patterns or outliers in real-time data as it is continuously generated. Unlike traditional methods that analyze historical data, streaming anomaly detection operates on live data streams, making it ideal for monitoring systems that require immediate detection and response to anomalies.

In IoT/OT environments, it's used to detect issues like equipment malfunctions, security breaches, or operational problems as they occur, enabling rapid intervention to prevent potential problems or mitigate their impact.

How does it work?

This detection mode is set from the events module and analyses each new data point as it is received. For more detail on setting the streaming anomoly mode refer to the events module.

Identifies if the current point of the time series data is anomaly.

This sends the most recent 1000 data points to the AlphaX intelligence engine to build a model, then analyzes the last data point based on that model.

Parameters

Feature	Description
Sensitivity	A numerical value (0% - 99%) used to control the tolerance of Anomaly Detection. Lower sensitivity results in fewer anomalies being detected. It is recommended to set the sensitivity below 75% to prevent over-sensitivity and reduce false positives.
Max Anomaly Ratio	Not used

Output

Feature	Description
Anomaly	Boolean: If the Last data point is Identified as an anomaly