Azure AI Anomaly Detector for Java FAQs

Question 1

How does the skill handle Azure authentication?

Accepted Answer

The skill provides patterns for both simple API key-based authentication and DefaultAzureCredential for secure, identity-based access management.

Question 2

Does this skill support real-time anomaly detection?

Accepted Answer

Yes, it includes implementation patterns for Univariate Last Point Detection (streaming) to identify anomalies instantly as new data points arrive.

Question 3

What is the difference between univariate and multivariate detection?

Accepted Answer

Univariate detection focuses on identifying anomalies within a single time-series signal, while multivariate detection identifies issues across multiple correlated signals simultaneously.

Question 4

Can I detect changes in data trends using this skill?

Accepted Answer

Yes, it includes dedicated patterns for Change Point Detection, which identifies when a time series undergoes a significant structural or trend shift.

Question 5

What is the minimum data requirement for detection?

Accepted Answer

Univariate detection requires a minimum of 12 data points, although providing more historical data significantly improves the accuracy and sensitivity of the results.

Azure AI Anomaly Detector for Java

Key Features

Use Cases

Azure AI Anomaly Detector for Java

Key Features

Use Cases