Log Service - Supports Anomaly Detection for Time Series Data Streams
Target customers: Users who use Log Service for intelligent analysis and operations and maintenance. Features released: 1. Log Service supports intelligent detection of structured time series data. You do not need to set the anomaly type for metric data. The backend algorithms identify the type of anomalies of metric data points and return anomaly detection results of the standard format. This makes it easier for you to configure alerts. 2.The tree-based algorithm for the anomaly detection of streaming data uses the Time2Graph model to analyze input time series curves. The time series data is converted to graphs. The anomalies of data points are determined by analyzing the anomaly transfer paths in the graphs.