Does it support multivariate time series data?

Yes, Aeon is designed to handle both univariate and multivariate time series data, typically expecting a shape of (n_samples, n_channels, n_timepoints).

Is Aeon compatible with scikit-learn?

Yes, Aeon is built to be scikit-learn compatible, allowing you to use its specialized time series estimators within standard pipelines and cross-validation workflows.

What is the best algorithm for fast time series classification?

For fast prototyping and high performance, the skill recommends using ROCKET variants like MiniRocketClassifier or Arsenal.

What is the Aeon Claude Code Skill?

It is a specialized extension for Claude Code that provides expert guidance, implementation patterns, and code snippets for the Aeon time series machine learning toolkit.

Can I use this skill for deep learning tasks?

Yes, this skill includes support for various neural architectures specifically designed for time series, such as InceptionTime, ResNet, and FCN.

Aeon Time Series Machine Learning

Name: Aeon Time Series Machine Learning
Author: ricable

byricable

•

数据科学与机器学习

Implements advanced time series machine learning workflows for classification, forecasting, and anomaly detection using scikit-learn compatible APIs.

关于

Aeon is a comprehensive Python toolkit designed for specialized time series machine learning tasks that go beyond standard machine learning approaches. It offers state-of-the-art algorithms for classification, regression, clustering, forecasting, and similarity search, all within a scikit-learn compatible framework. Whether you are analyzing univariate or multivariate temporal data, this skill provides the patterns and best practices needed to extract insights from sequential patterns, detect anomalies, and predict future trends using both traditional and deep learning architectures.

主要功能

Advanced forecasting models including ARIMA and specialized neural networks
Extensive library of temporal distance metrics including Dynamic Time Warping (DTW)
Comprehensive suite for time series classification and regression using ROCKET and HIVE-COTE
Deep learning architectures like InceptionTime optimized for temporal sequences
Specialized anomaly detection and segmentation for identifying temporal change points
1 GitHub stars

使用场景

Classifying movement patterns or rhythmic signals in medical and IoT telemetry
Detecting anomalies and change points in high-frequency industrial sensor data
Forecasting future values for demand planning and financial trend analysis

关于

主要功能

Advanced forecasting models including ARIMA and specialized neural networks
Extensive library of temporal distance metrics including Dynamic Time Warping (DTW)
Comprehensive suite for time series classification and regression using ROCKET and HIVE-COTE
Deep learning architectures like InceptionTime optimized for temporal sequences
Specialized anomaly detection and segmentation for identifying temporal change points
1 GitHub stars

使用场景

Classifying movement patterns or rhythmic signals in medical and IoT telemetry
Detecting anomalies and change points in high-frequency industrial sensor data
Forecasting future values for demand planning and financial trend analysis