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

•

Data Science & ML

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

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.

Key Features

01Advanced forecasting models including ARIMA and specialized neural networks

02Extensive library of temporal distance metrics including Dynamic Time Warping (DTW)

03Comprehensive suite for time series classification and regression using ROCKET and HIVE-COTE

04Deep learning architectures like InceptionTime optimized for temporal sequences

05Specialized anomaly detection and segmentation for identifying temporal change points

061 GitHub stars

Use Cases

01Classifying movement patterns or rhythmic signals in medical and IoT telemetry

02Detecting anomalies and change points in high-frequency industrial sensor data

03Forecasting future values for demand planning and financial trend analysis

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add ricable/ultimate-ai-agent aeon

For use in Claude.ai and ChatGPT

Download Skill