Aeon Time Series Toolkit

关于

Aeon is a unified Python toolkit for time series machine learning, offering a vast array of algorithms for classification, regression, clustering, forecasting, and more. It bridges the gap between classical statistical techniques and modern deep learning, providing specialized tools for distance metrics (DTW), feature extraction (ROCKET), and temporal segmentation. Whether you are analyzing financial markets, interpreting sensor data, or building predictive maintenance systems, Aeon provides the standardized framework and state-of-the-art models needed to extract meaningful insights from sequential data.

主要功能

• Specialized distance metrics like Dynamic Time Warping (DTW) and MSM
• 16 GitHub stars
• Deep learning architectures optimized for sequential data like InceptionTime
• Advanced feature extraction using ROCKET, Catch22, and TSFresh
• Standardized workflows for temporal segmentation and anomaly detection
• Seven primary temporal learning tasks including classification and forecasting

使用场景

• Classifying temporal patterns in medical ECG or wearable activity data
• Identifying anomalies and motifs in real-time sensor streams
• Forecasting future values in financial markets or energy demand