aeon

1# Aeon Time Series Machine Learning
2 
3## Overview
4 
5Aeon is a scikit-learn compatible Python toolkit for time series machine learning. It provides state-of-the-art algorithms for classification, regression, clustering, forecasting, anomaly detection, segmentation, and similarity search.
6 
7## When to Use This Skill
8 
9Apply this skill when:
10- Classifying or predicting from time series data
11- Detecting anomalies or change points in temporal sequences
12- Clustering similar time series patterns
13- Forecasting future values
14- Finding repeated patterns (motifs) or unusual subsequences (discords)
15- Comparing time series with specialized distance metrics
16- Extracting features from temporal data
17 
18## Installation
19 
20```bash
21uv pip install aeon
22```
23 
24## Core Capabilities
25 
26### 1. Time Series Classification
27 
28Categorize time series into predefined classes. See `references/classification.md` for complete algorithm catalog.
29 
30**Quick Start:**
31```python
32from aeon.classification.convolution_based import RocketClassifier
33from aeon.datasets import load_classification
34 
35# Load data
36X_train, y_train = load_classification("GunPoint", split="train")
37X_test, y_test = load_classification("GunPoint", split="test")
38 
39# Train classifier
40clf = RocketClassifier(n_kernels=10000)
41clf.fit(X_train, y_train)
42accuracy = clf.score(X_test, y_test)
43```
44 
45**Algorithm Selection:**
46- **Speed + Performance**: `MiniRocketClassifier`, `Arsenal`
47- **Maximum Accuracy**: `HIVECOTEV2`, `InceptionTimeClassifier`
48…