Anomaly Detection as a Service - What is normal?
summary
The talk explains how to detect anomalies in time-based application metrics (like purchases, website accesses, sent messages, or errors) using Azure Cognitive Services' Metrics Advisor. It demonstrates, with live demos and practical examples, how to call the Metrics Advisor APIs to get AI-assisted anomaly detection for those time-series datasets.
AI-written summary · about
chapters (10)
- 0:00Welcome
- 1:16What is anomaly detection?
- 4:05Time-based anomaly detection
- 10:26Magic Advisor overview
- 16:05Data sources and granularity
- 20:35Demo: Setting up data feed
- 28:40Technical background
- 34:17Live demo: Application Insights
- 44:21Notifications and feedback
- 48:10Summary
related talks
45mAnomaly Detection in Time Series
38mAnomaly detection in the cloud: how to detect issues in data [...]
19mThe Rise of AI in Anomaly Detection
21mCross-dataset Time Series Anomaly Detection for Cloud Systems
28mReal Time Sensor Anomaly Detection with Sci Kit Learn and the Azure Stack
35mObservability’s Sixth Sense: Detecting Anomalies in Metrics
22mAnomaly Detection: The Bad Boy of Data Science
39mDeveloping Spidey Senses