Anomaly detection in streaming time series data with online learning using Amazon Managed Service for Apache Flink
Machine Learning Blog
This article discusses a solution for real-time anomaly detection in streaming time series data using Amazon Managed Service for Apache Flink and other AWS services.
Specifically, the article covers:
- Overview of the Anomaly Detection Stack solution architecture
- Ingestion layer using AWS Lambda, Amazon EventBridge, and Amazon MSK
- Anomaly detection stack using Flink and Amazon SageMaker
- Transform layer using Amazon Firehose, AWS Lambda, Amazon S3, AWS Glue, and Amazon Athena
- Visualization using Amazon QuickSight dashboard
- Notification layer using AWS Lambda and Amazon SNS
- Conclusion highlighting the benefits of the solution
The AWS News Feed is currently looking for gold sponsors. If you want to support the AWS community and reach a large audience of AWS professionals, consider sponsoring the AWS News Feed.
Related articles
Aug 8
2024
2024
Introducing AWS Glue Data Quality anomaly detection
Feb 1
2024
2024
Build a real-time, low-code anomaly detection pipeline for time series data using Amazon Aurora, Amazon Redshift ML, and Amazon SageMaker
Mar 8
2024
2024
In-stream anomaly detection with Amazon OpenSearch Ingestion and Amazon OpenSearch Serverless
Feb 16
2024
2024
Predictive Analytics with Time-series Machine Learning on Amazon Timestream
The AWS News Feed is currently looking for silver sponsors. If you want to support the AWS community and reach a large audience of AWS professionals, consider sponsoring the AWS News Feed.