aeon

# Aeon Time Series Machine Learning

## Overview

Aeon 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.

## When to Use This Skill

Apply this skill when:
- Classifying or predicting from time series data
- Detecting anomalies or change points in temporal sequences
- Clustering similar time series patterns
- Forecasting future values
- Finding repeated patterns (motifs) or unusual subsequences (discords)
- Comparing time series with specialized distance metrics
- Extracting features from temporal data

## Installation

```bash
uv pip install aeon
```

## Core Capabilities

### 1. Time Series Classification

Categorize time series into predefined classes. See `references/classification.md` for complete algorithm catalog.

**Quick Start:**
```python
from aeon.classification.convolution_based import RocketClassifier
from aeon.datasets import load_classification

# Load data
X_train, y_train = load_classification("GunPoint", split="train")
X_test, y_test = load_classification("GunPoint", split="test")

# Train classifier
clf = RocketClassifier(n_kernels=10000)
clf.fit(X_train, y_train)
accuracy = clf.score(X_test, y_test)
```

**Algorithm Selection:**
- **Speed + Performance**: `MiniRocketClassifier`, `Arsenal`
- **Maximum Accuracy**: `HIVECOTEV2`, `InceptionTimeClassifier`
- **Interpretability**: `ShapeletTransformClassifier`, `Catch22Classifier`
- **Small Datasets**: `KNeighborsTimeSeriesClassifier` with DTW distance

### 2. Time Series Regression

Predict continuous values from time series. See `references/regression.md` for algorithms.

**Quick Start:**
```python
from aeon.regression.convolution_based import RocketRegressor
from aeon.datasets import load_regression

X_train, y_train = load_regression("Covid3Month", split="train")
X_test, y_test = load_regression("Covid3Month", split="test")

reg = RocketRegressor()
reg.fit(X_train, y_train)
predictions = reg.predict(X_test)
```

### 3. Time Series Clustering

Group similar time series without labels. See `references/clustering.md` for methods.

**Quick Start:**
```python
from aeon.clustering import TimeSeriesKMeans

clusterer = TimeSeriesKMeans(
    n_clusters=3,
    distance="dtw",
    averaging_method="ba"
)
labels = clusterer.fit_predict(X_train)
centers = clusterer.cluster_centers_
```

### 4. Forecasting

Predict future time series values. See `references/forecasting.md` for forecasters.

**Quick Start:**
```python
from aeon.forecasting.arima import ARIMA

forecaster = ARIMA(order=(1, 1, 1))
forecaster.fit(y_train)
y_pred = forecaster.predict(fh=[1, 2, 3, 4, 5])
```

### 5. Anomaly Detection

Identify unusual patterns or outliers. See `references/anomaly_detection.md` for detectors.

**Quick Start:**
```python
from aeon.anomaly_detection import STOMP

detector = STOMP(window_size=50)
anomaly_scores = detector.fit_predict(y)

# Higher scores indicate anomalies
threshold = np.percentile(anomaly_scores, 95)
anomalies = anomaly_scores > threshold
```

### 6. Segmentation

Partition time series into regions with change points. See `references/segmentation.md`.

**Quick Start:**
```python
from aeon.segmentation import ClaSPSegmenter

segmenter = ClaSPSegmenter()
change_points = segmenter.fit_predict(y)
```

### 7. Similarity Search

Find similar patterns within or across time series. See `references/similarity_search.md`.

**Quick Start:**
```python
from aeon.similarity_search import StompMotif

# Find recurring patterns
motif_finder = StompMotif(window_size=50, k=3)
motifs = motif_finder.fit_predict(y)
```

## Feature Extraction and Transformations

Transform time series for feature engineering. See `references/transformations.md`.

**ROCKET Features:**
```python
from aeon.transformations.collection.convolution_based import RocketTransformer

rocket = RocketTransformer()
X_features = rocket.fit_transform(X_train)

# Use features with any sklearn classifier
from sklearn.ensemble import RandomForestClassifier
clf = RandomForestClassifier()
clf.fit(X_features, y_train)
```

**Statistical Features:**
```python
from aeon.transformations.collection.feature_based import Catch22

catch22 = Catch22()
X_features = catch22.fit_transform(X_train)
```

**Preprocessing:**
```python
from aeon.transformations.collection import MinMaxScaler, Normalizer

scaler = Normalizer()  # Z-normalization
X_normalized = scaler.fit_transform(X_train)
```

## Distance Metrics

Specialized temporal distance measures. See `references/distances.md` for complete catalog.

**Usage:**
```python
from aeon.distances import dtw_distance, dtw_pairwise_distance

# Single distance
distance = dtw_distance(x, y, window=0.1)

# Pairwise distances
distance_matrix = dtw_pairwise_distance(X_train)

# Use with classifiers
from aeon.classification.distance_based import KNeighborsTimeSeriesClassifier

clf = KNeighborsTimeSeriesClassifier(
    n_neighbors=5,
    distance="dtw",
    distance_params={"window": 0.2}
)
```

**Available Distances:**
- **Elastic**: DTW, DDTW, WDTW, ERP, EDR, LCSS, TWE, MSM
- **Lock-step**: Euclidean, Manhattan, Minkowski
- **Shape-based**: Shape DTW, SBD

## Deep Learning Networks

Neural architectures for time series. See `references/networks.md`.

**Architectures:**
- Convolutional: `FCNClassifier`, `ResNetClassifier`, `InceptionTimeClassifier`
- Recurrent: `RecurrentNetwork`, `TCNNetwork`
- Autoencoders: `AEFCNClusterer`, `AEResNetClusterer`

**Usage:**
```python
from aeon.classification.deep_learning import InceptionTimeClassifier

clf = InceptionTimeClassifier(n_epochs=100, batch_size=32)
clf.fit(X_train, y_train)
predictions = clf.predict(X_test)
```

## Datasets and Benchmarking

Load standard benchmarks and evaluate performance. See `references/datasets_benchmarking.md`.

**Load Datasets:**
```python
from aeon.datasets import load_classification, load_regression

# Classification
X_train, y_train = load_classification("ArrowHead", split="train")

# Regression
X_train, y_train = load_regression("Covid3Month", split="train")
```

**Benchmarking:**
```python
from aeon.benchmarking import get_estimator_results

# Compare with published results
published = get_estimator_results("ROCKET", "GunPoint")
```

## Common Workflows

### Classification Pipeline

```python
from aeon.transformations.collection import Normalizer
from aeon.classification.convolution_based import RocketClassifier
from sklearn.pipeline import Pipeline

pipeline = Pipeline([
    ('normalize', Normalizer()),
    ('classify', RocketClassifier())
])

pipeline.fit(X_train, y_train)
accuracy = pipeline.score(X_test, y_test)
```

### Feature Extraction + Traditional ML

```python
from aeon.transformations.collection import RocketTransformer
from sklearn.ensemble import GradientBoostingClassifier

# Extract features
rocket = RocketTransformer()
X_train_features = rocket.fit_transform(X_train)
X_test_features = rocket.transform(X_test)

# Train traditional ML
clf = GradientBoostingClassifier()
clf.fit(X_train_features, y_train)
predictions = clf.predict(X_test_features)
```

### Anomaly Detection with Visualization

```python
from aeon.anomaly_detection import STOMP
import matplotlib.pyplot as plt

detector = STOMP(window_size=50)
scores = detector.fit_predict(y)

plt.figure(figsize=(15, 5))
plt.subplot(2, 1, 1)
plt.plot(y, label='Time Series')
plt.subplot(2, 1, 2)
plt.plot(scores, label='Anomaly Scores', color='red')
plt.axhline(np.percentile(scores, 95), color='k', linestyle='--')
plt.show()
```

## Best Practices

### Data Preparation

1. **Normalize**: Most algorithms benefit from z-normalization
   ```python
   from aeon.transformations.collection import Normalizer
   normalizer = Normalizer()
   X_train = normalizer.fit_transform(X_train)
   X_test = normalizer.transform(X_test)
   ```

2. **Handle Missing Values**: Impute before analysis
   ```python
   from aeon.transformations.collection import SimpleImputer
   imputer = SimpleImputer(strategy='mean')
   X_train = imputer.fit_transform(X_train)
   ```

3. **Check Data Format**: Aeon expects shape `(n_samples, n_channels, n_timepoints)`

### Model Selection

1. **Start Simple**: Begin with ROCKET variants before deep learning
2. **Use Validation**: Split training data for hyperparameter tuning
3. **Compare Baselines**: Test against simple methods (1-NN Euclidean, Naive)
4. **Consider Resources**: ROCKET for speed, deep learning if GPU available

### Algorithm Selection Guide

**For Fast Prototyping:**
- Classification: `MiniRocketClassifier`
- Regression: `MiniRocketRegressor`
- Clustering: `TimeSeriesKMeans` with Euclidean

**For Maximum Accuracy:**
- Classification: `HIVECOTEV2`, `InceptionTimeClassifier`
- Regression: `InceptionTimeRegressor`
- Forecasting: `ARIMA`, `TCNForecaster`

**For Interpretability:**
- Classification: `ShapeletTransformClassifier`, `Catch22Classifier`
- Features: `Catch22`, `TSFresh`

**For Small Datasets:**
- Distance-based: `KNeighborsTimeSeriesClassifier` with DTW
- Avoid: Deep learning (requires large data)

## Reference Documentation

Detailed information available in `references/`:
- `classification.md` - All classification algorithms
- `regression.md` - Regression methods
- `clustering.md` - Clustering algorithms
- `forecasting.md` - Forecasting approaches
- `anomaly_detection.md` - Anomaly detection methods
- `segmentation.md` - Segmentation algorithms
- `similarity_search.md` - Pattern matching and motif discovery
- `transformations.md` - Feature extraction and preprocessing
- `distances.md` - Time series distance metrics
- `networks.md` - Deep learning architectures
- `datasets_benchmarking.md` - Data loading and evaluation tools

## Additional Resources

- Documentation: https://www.aeon-toolkit.org/
- GitHub: https://github.com/aeon-toolkit/aeon
- Examples: https://www.aeon-toolkit.org/en/stable/examples.html
- API Reference: https://www.aeon-toolkit.org/en/stable/api_reference.html

{
  "format": "aiskill/v1",
  "skill": {
    "name": "aeon",
    "slug": "aeon",
    "version": "1.0.0",
    "description": "This skill should be used for time series machine learning tasks including classification, regression, clustering, forecasting, anomaly detection, segmentation, and similarity search. Use when working with temporal data, sequential patterns, or time-indexed observations requiring specialized algorithms beyond standard ML approaches. Particularly suited for univariate and multivariate time series analysis with scikit-learn compatible APIs.",
    "category": "Research",
    "tags": [
      "research-science"
    ],
    "platforms": [
      "chatgpt",
      "claude-code",
      "cursor"
    ],
    "tested_environment": "Agent Skills convention (SKILL.md) — works with Claude Code, Cursor, ChatGPT/Codex and compatible agents",
    "license": "BSD-3-CLAUSE LICENSE"
  },
  "prompts": {
    "system": "# Aeon Time Series Machine Learning\r\n\r\n## Overview\r\n\r\nAeon 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.\r\n\r\n## When to Use This Skill\r\n\r\nApply this skill when:\r\n- Classifying or predicting from time series data\r\n- Detecting anomalies or change points in temporal sequences\r\n- Clustering similar time series patterns\r\n- Forecasting future values\r\n- Finding repeated patterns (motifs) or unusual subsequences (discords)\r\n- Comparing time series with specialized distance metrics\r\n- Extracting features from temporal data\r\n\r\n## Installation\r\n\r\n```bash\r\nuv pip install aeon\r\n```\r\n\r\n## Core Capabilities\r\n\r\n### 1. Time Series Classification\r\n\r\nCategorize time series into predefined classes. See `references/classification.md` for complete algorithm catalog.\r\n\r\n**Quick Start:**\r\n```python\r\nfrom aeon.classification.convolution_based import RocketClassifier\r\nfrom aeon.datasets import load_classification\r\n\r\n# Load data\r\nX_train, y_train = load_classification(\"GunPoint\", split=\"train\")\r\nX_test, y_test = load_classification(\"GunPoint\", split=\"test\")\r\n\r\n# Train classifier\r\nclf = RocketClassifier(n_kernels=10000)\r\nclf.fit(X_train, y_train)\r\naccuracy = clf.score(X_test, y_test)\r\n```\r\n\r\n**Algorithm Selection:**\r\n- **Speed + Performance**: `MiniRocketClassifier`, `Arsenal`\r\n- **Maximum Accuracy**: `HIVECOTEV2`, `InceptionTimeClassifier`\r\n- **Interpretability**: `ShapeletTransformClassifier`, `Catch22Classifier`\r\n- **Small Datasets**: `KNeighborsTimeSeriesClassifier` with DTW distance\r\n\r\n### 2. Time Series Regression\r\n\r\nPredict continuous values from time series. See `references/regression.md` for algorithms.\r\n\r\n**Quick Start:**\r\n```python\r\nfrom aeon.regression.convolution_based import RocketRegressor\r\nfrom aeon.datasets import load_regression\r\n\r\nX_train, y_train = load_regression(\"Covid3Month\", split=\"train\")\r\nX_test, y_test = load_regression(\"Covid3Month\", split=\"test\")\r\n\r\nreg = RocketRegressor()\r\nreg.fit(X_train, y_train)\r\npredictions = reg.predict(X_test)\r\n```\r\n\r\n### 3. Time Series Clustering\r\n\r\nGroup similar time series without labels. See `references/clustering.md` for methods.\r\n\r\n**Quick Start:**\r\n```python\r\nfrom aeon.clustering import TimeSeriesKMeans\r\n\r\nclusterer = TimeSeriesKMeans(\r\n    n_clusters=3,\r\n    distance=\"dtw\",\r\n    averaging_method=\"ba\"\r\n)\r\nlabels = clusterer.fit_predict(X_train)\r\ncenters = clusterer.cluster_centers_\r\n```\r\n\r\n### 4. Forecasting\r\n\r\nPredict future time series values. See `references/forecasting.md` for forecasters.\r\n\r\n**Quick Start:**\r\n```python\r\nfrom aeon.forecasting.arima import ARIMA\r\n\r\nforecaster = ARIMA(order=(1, 1, 1))\r\nforecaster.fit(y_train)\r\ny_pred = forecaster.predict(fh=[1, 2, 3, 4, 5])\r\n```\r\n\r\n### 5. Anomaly Detection\r\n\r\nIdentify unusual patterns or outliers. See `references/anomaly_detection.md` for detectors.\r\n\r\n**Quick Start:**\r\n```python\r\nfrom aeon.anomaly_detection import STOMP\r\n\r\ndetector = STOMP(window_size=50)\r\nanomaly_scores = detector.fit_predict(y)\r\n\r\n# Higher scores indicate anomalies\r\nthreshold = np.percentile(anomaly_scores, 95)\r\nanomalies = anomaly_scores > threshold\r\n```\r\n\r\n### 6. Segmentation\r\n\r\nPartition time series into regions with change points. See `references/segmentation.md`.\r\n\r\n**Quick Start:**\r\n```python\r\nfrom aeon.segmentation import ClaSPSegmenter\r\n\r\nsegmenter = ClaSPSegmenter()\r\nchange_points = segmenter.fit_predict(y)\r\n```\r\n\r\n### 7. Similarity Search\r\n\r\nFind similar patterns within or across time series. See `references/similarity_search.md`.\r\n\r\n**Quick Start:**\r\n```python\r\nfrom aeon.similarity_search import StompMotif\r\n\r\n# Find recurring patterns\r\nmotif_finder = StompMotif(window_size=50, k=3)\r\nmotifs = motif_finder.fit_predict(y)\r\n```\r\n\r\n## Feature Extraction and Transformations\r\n\r\nTransform time series for feature engineering. See `references/transformations.md`.\r\n\r\n**ROCKET Features:**\r\n```python\r\nfrom aeon.transformations.collection.convolution_based import RocketTransformer\r\n\r\nrocket = RocketTransformer()\r\nX_features = rocket.fit_transform(X_train)\r\n\r\n# Use features with any sklearn classifier\r\nfrom sklearn.ensemble import RandomForestClassifier\r\nclf = RandomForestClassifier()\r\nclf.fit(X_features, y_train)\r\n```\r\n\r\n**Statistical Features:**\r\n```python\r\nfrom aeon.transformations.collection.feature_based import Catch22\r\n\r\ncatch22 = Catch22()\r\nX_features = catch22.fit_transform(X_train)\r\n```\r\n\r\n**Preprocessing:**\r\n```python\r\nfrom aeon.transformations.collection import MinMaxScaler, Normalizer\r\n\r\nscaler = Normalizer()  # Z-normalization\r\nX_normalized = scaler.fit_transform(X_train)\r\n```\r\n\r\n## Distance Metrics\r\n\r\nSpecialized temporal distance measures. See `references/distances.md` for complete catalog.\r\n\r\n**Usage:**\r\n```python\r\nfrom aeon.distances import dtw_distance, dtw_pairwise_distance\r\n\r\n# Single distance\r\ndistance = dtw_distance(x, y, window=0.1)\r\n\r\n# Pairwise distances\r\ndistance_matrix = dtw_pairwise_distance(X_train)\r\n\r\n# Use with classifiers\r\nfrom aeon.classification.distance_based import KNeighborsTimeSeriesClassifier\r\n\r\nclf = KNeighborsTimeSeriesClassifier(\r\n    n_neighbors=5,\r\n    distance=\"dtw\",\r\n    distance_params={\"window\": 0.2}\r\n)\r\n```\r\n\r\n**Available Distances:**\r\n- **Elastic**: DTW, DDTW, WDTW, ERP, EDR, LCSS, TWE, MSM\r\n- **Lock-step**: Euclidean, Manhattan, Minkowski\r\n- **Shape-based**: Shape DTW, SBD\r\n\r\n## Deep Learning Networks\r\n\r\nNeural architectures for time series. See `references/networks.md`.\r\n\r\n**Architectures:**\r\n- Convolutional: `FCNClassifier`, `ResNetClassifier`, `InceptionTimeClassifier`\r\n- Recurrent: `RecurrentNetwork`, `TCNNetwork`\r\n- Autoencoders: `AEFCNClusterer`, `AEResNetClusterer`\r\n\r\n**Usage:**\r\n```python\r\nfrom aeon.classification.deep_learning import InceptionTimeClassifier\r\n\r\nclf = InceptionTimeClassifier(n_epochs=100, batch_size=32)\r\nclf.fit(X_train, y_train)\r\npredictions = clf.predict(X_test)\r\n```\r\n\r\n## Datasets and Benchmarking\r\n\r\nLoad standard benchmarks and evaluate performance. See `references/datasets_benchmarking.md`.\r\n\r\n**Load Datasets:**\r\n```python\r\nfrom aeon.datasets import load_classification, load_regression\r\n\r\n# Classification\r\nX_train, y_train = load_classification(\"ArrowHead\", split=\"train\")\r\n\r\n# Regression\r\nX_train, y_train = load_regression(\"Covid3Month\", split=\"train\")\r\n```\r\n\r\n**Benchmarking:**\r\n```python\r\nfrom aeon.benchmarking import get_estimator_results\r\n\r\n# Compare with published results\r\npublished = get_estimator_results(\"ROCKET\", \"GunPoint\")\r\n```\r\n\r\n## Common Workflows\r\n\r\n### Classification Pipeline\r\n\r\n```python\r\nfrom aeon.transformations.collection import Normalizer\r\nfrom aeon.classification.convolution_based import RocketClassifier\r\nfrom sklearn.pipeline import Pipeline\r\n\r\npipeline = Pipeline([\r\n    ('normalize', Normalizer()),\r\n    ('classify', RocketClassifier())\r\n])\r\n\r\npipeline.fit(X_train, y_train)\r\naccuracy = pipeline.score(X_test, y_test)\r\n```\r\n\r\n### Feature Extraction + Traditional ML\r\n\r\n```python\r\nfrom aeon.transformations.collection import RocketTransformer\r\nfrom sklearn.ensemble import GradientBoostingClassifier\r\n\r\n# Extract features\r\nrocket = RocketTransformer()\r\nX_train_features = rocket.fit_transform(X_train)\r\nX_test_features = rocket.transform(X_test)\r\n\r\n# Train traditional ML\r\nclf = GradientBoostingClassifier()\r\nclf.fit(X_train_features, y_train)\r\npredictions = clf.predict(X_test_features)\r\n```\r\n\r\n### Anomaly Detection with Visualization\r\n\r\n```python\r\nfrom aeon.anomaly_detection import STOMP\r\nimport matplotlib.pyplot as plt\r\n\r\ndetector = STOMP(window_size=50)\r\nscores = detector.fit_predict(y)\r\n\r\nplt.figure(figsize=(15, 5))\r\nplt.subplot(2, 1, 1)\r\nplt.plot(y, label='Time Series')\r\nplt.subplot(2, 1, 2)\r\nplt.plot(scores, label='Anomaly Scores', color='red')\r\nplt.axhline(np.percentile(scores, 95), color='k', linestyle='--')\r\nplt.show()\r\n```\r\n\r\n## Best Practices\r\n\r\n### Data Preparation\r\n\r\n1. **Normalize**: Most algorithms benefit from z-normalization\r\n   ```python\r\n   from aeon.transformations.collection import Normalizer\r\n   normalizer = Normalizer()\r\n   X_train = normalizer.fit_transform(X_train)\r\n   X_test = normalizer.transform(X_test)\r\n   ```\r\n\r\n2. **Handle Missing Values**: Impute before analysis\r\n   ```python\r\n   from aeon.transformations.collection import SimpleImputer\r\n   imputer = SimpleImputer(strategy='mean')\r\n   X_train = imputer.fit_transform(X_train)\r\n   ```\r\n\r\n3. **Check Data Format**: Aeon expects shape `(n_samples, n_channels, n_timepoints)`\r\n\r\n### Model Selection\r\n\r\n1. **Start Simple**: Begin with ROCKET variants before deep learning\r\n2. **Use Validation**: Split training data for hyperparameter tuning\r\n3. **Compare Baselines**: Test against simple methods (1-NN Euclidean, Naive)\r\n4. **Consider Resources**: ROCKET for speed, deep learning if GPU available\r\n\r\n### Algorithm Selection Guide\r\n\r\n**For Fast Prototyping:**\r\n- Classification: `MiniRocketClassifier`\r\n- Regression: `MiniRocketRegressor`\r\n- Clustering: `TimeSeriesKMeans` with Euclidean\r\n\r\n**For Maximum Accuracy:**\r\n- Classification: `HIVECOTEV2`, `InceptionTimeClassifier`\r\n- Regression: `InceptionTimeRegressor`\r\n- Forecasting: `ARIMA`, `TCNForecaster`\r\n\r\n**For Interpretability:**\r\n- Classification: `ShapeletTransformClassifier`, `Catch22Classifier`\r\n- Features: `Catch22`, `TSFresh`\r\n\r\n**For Small Datasets:**\r\n- Distance-based: `KNeighborsTimeSeriesClassifier` with DTW\r\n- Avoid: Deep learning (requires large data)\r\n\r\n## Reference Documentation\r\n\r\nDetailed information available in `references/`:\r\n- `classification.md` - All classification algorithms\r\n- `regression.md` - Regression methods\r\n- `clustering.md` - Clustering algorithms\r\n- `forecasting.md` - Forecasting approaches\r\n- `anomaly_detection.md` - Anomaly detection methods\r\n- `segmentation.md` - Segmentation algorithms\r\n- `similarity_search.md` - Pattern matching and motif discovery\r\n- `transformations.md` - Feature extraction and preprocessing\r\n- `distances.md` - Time series distance metrics\r\n- `networks.md` - Deep learning architectures\r\n- `datasets_benchmarking.md` - Data loading and evaluation tools\r\n\r\n## Additional Resources\r\n\r\n- Documentation: https://www.aeon-toolkit.org/\r\n- GitHub: https://github.com/aeon-toolkit/aeon\r\n- Examples: https://www.aeon-toolkit.org/en/stable/examples.html\r\n- API Reference: https://www.aeon-toolkit.org/en/stable/api_reference.html",
    "user_template": ""
  },
  "variables": [],
  "install": {
    "steps": [
      {
        "title": "Download & unzip",
        "detail": "Download the .zip package from SkillVault and unzip it. The aeon/ folder contains SKILL.md plus any companion scripts, references and assets."
      },
      {
        "title": "Copy into your agent skills directory",
        "detail": "Move the folder into your agent's skills directory — e.g. ~/.claude/skills/ (Claude Code), .codex/skills/, .cursor/skills/ or your project's .claude/skills/ folder — then restart or reload the session."
      },
      {
        "title": "Use it",
        "detail": "Describe the task in natural language; the agent loads the skill automatically when the request matches the skill description."
      }
    ]
  },
  "usage": {
    "steps": []
  },
  "metadata": {
    "author": "SkillVault",
    "source": "https://skillvault.top/skills/aeon",
    "exported_at": "2026-11-03T07:41:10.794Z"
  }
}

format: aiskill/v1
skill:
  name: aeon
  slug: aeon
  version: 1.0.0
  description: >-
    This skill should be used for time series machine learning tasks including classification,
    regression, clustering, forecasting, anomaly detection, segmentation, and similarity search. Use
    when working with temporal data, sequential patterns, or time-indexed observations requiring
    specialized algorithms beyond standard ML approaches. Particularly suited for univariate and
    multivariate time series analysis with scikit-learn compatible APIs.
  category: Research
  tags:
    - research-science
  platforms:
    - chatgpt
    - claude-code
    - cursor
  tested_environment: >-
    Agent Skills convention (SKILL.md) — works with Claude Code, Cursor, ChatGPT/Codex and
    compatible agents
  license: BSD-3-CLAUSE LICENSE
prompts:
  system: "# Aeon Time Series Machine Learning\r\n\r\n## Overview\r\n\r\nAeon 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.\r\n\r\n## When to Use This Skill\r\n\r\nApply this skill when:\r\n- Classifying or predicting from time series data\r\n- Detecting anomalies or change points in temporal sequences\r\n- Clustering similar time series patterns\r\n- Forecasting future values\r\n- Finding repeated patterns (motifs) or unusual subsequences (discords)\r\n- Comparing time series with specialized distance metrics\r\n- Extracting features from temporal data\r\n\r\n## Installation\r\n\r\n```bash\r\nuv pip install aeon\r\n```\r\n\r\n## Core Capabilities\r\n\r\n### 1. Time Series Classification\r\n\r\nCategorize time series into predefined classes. See `references/classification.md` for complete algorithm catalog.\r\n\r\n**Quick Start:**\r\n```python\r\nfrom aeon.classification.convolution_based import RocketClassifier\r\nfrom aeon.datasets import load_classification\r\n\r\n# Load data\r\nX_train, y_train = load_classification(\"GunPoint\", split=\"train\")\r\nX_test, y_test = load_classification(\"GunPoint\", split=\"test\")\r\n\r\n# Train classifier\r\nclf = RocketClassifier(n_kernels=10000)\r\nclf.fit(X_train, y_train)\r\naccuracy = clf.score(X_test, y_test)\r\n```\r\n\r\n**Algorithm Selection:**\r\n- **Speed + Performance**: `MiniRocketClassifier`, `Arsenal`\r\n- **Maximum Accuracy**: `HIVECOTEV2`, `InceptionTimeClassifier`\r\n- **Interpretability**: `ShapeletTransformClassifier`, `Catch22Classifier`\r\n- **Small Datasets**: `KNeighborsTimeSeriesClassifier` with DTW distance\r\n\r\n### 2. Time Series Regression\r\n\r\nPredict continuous values from time series. See `references/regression.md` for algorithms.\r\n\r\n**Quick Start:**\r\n```python\r\nfrom aeon.regression.convolution_based import RocketRegressor\r\nfrom aeon.datasets import load_regression\r\n\r\nX_train, y_train = load_regression(\"Covid3Month\", split=\"train\")\r\nX_test, y_test = load_regression(\"Covid3Month\", split=\"test\")\r\n\r\nreg = RocketRegressor()\r\nreg.fit(X_train, y_train)\r\npredictions = reg.predict(X_test)\r\n```\r\n\r\n### 3. Time Series Clustering\r\n\r\nGroup similar time series without labels. See `references/clustering.md` for methods.\r\n\r\n**Quick Start:**\r\n```python\r\nfrom aeon.clustering import TimeSeriesKMeans\r\n\r\nclusterer = TimeSeriesKMeans(\r\n    n_clusters=3,\r\n    distance=\"dtw\",\r\n    averaging_method=\"ba\"\r\n)\r\nlabels = clusterer.fit_predict(X_train)\r\ncenters = clusterer.cluster_centers_\r\n```\r\n\r\n### 4. Forecasting\r\n\r\nPredict future time series values. See `references/forecasting.md` for forecasters.\r\n\r\n**Quick Start:**\r\n```python\r\nfrom aeon.forecasting.arima import ARIMA\r\n\r\nforecaster = ARIMA(order=(1, 1, 1))\r\nforecaster.fit(y_train)\r\ny_pred = forecaster.predict(fh=[1, 2, 3, 4, 5])\r\n```\r\n\r\n### 5. Anomaly Detection\r\n\r\nIdentify unusual patterns or outliers. See `references/anomaly_detection.md` for detectors.\r\n\r\n**Quick Start:**\r\n```python\r\nfrom aeon.anomaly_detection import STOMP\r\n\r\ndetector = STOMP(window_size=50)\r\nanomaly_scores = detector.fit_predict(y)\r\n\r\n# Higher scores indicate anomalies\r\nthreshold = np.percentile(anomaly_scores, 95)\r\nanomalies = anomaly_scores > threshold\r\n```\r\n\r\n### 6. Segmentation\r\n\r\nPartition time series into regions with change points. See `references/segmentation.md`.\r\n\r\n**Quick Start:**\r\n```python\r\nfrom aeon.segmentation import ClaSPSegmenter\r\n\r\nsegmenter = ClaSPSegmenter()\r\nchange_points = segmenter.fit_predict(y)\r\n```\r\n\r\n### 7. Similarity Search\r\n\r\nFind similar patterns within or across time series. See `references/similarity_search.md`.\r\n\r\n**Quick Start:**\r\n```python\r\nfrom aeon.similarity_search import StompMotif\r\n\r\n# Find recurring patterns\r\nmotif_finder = StompMotif(window_size=50, k=3)\r\nmotifs = motif_finder.fit_predict(y)\r\n```\r\n\r\n## Feature Extraction and Transformations\r\n\r\nTransform time series for feature engineering. See `references/transformations.md`.\r\n\r\n**ROCKET Features:**\r\n```python\r\nfrom aeon.transformations.collection.convolution_based import RocketTransformer\r\n\r\nrocket = RocketTransformer()\r\nX_features = rocket.fit_transform(X_train)\r\n\r\n# Use features with any sklearn classifier\r\nfrom sklearn.ensemble import RandomForestClassifier\r\nclf = RandomForestClassifier()\r\nclf.fit(X_features, y_train)\r\n```\r\n\r\n**Statistical Features:**\r\n```python\r\nfrom aeon.transformations.collection.feature_based import Catch22\r\n\r\ncatch22 = Catch22()\r\nX_features = catch22.fit_transform(X_train)\r\n```\r\n\r\n**Preprocessing:**\r\n```python\r\nfrom aeon.transformations.collection import MinMaxScaler, Normalizer\r\n\r\nscaler = Normalizer()  # Z-normalization\r\nX_normalized = scaler.fit_transform(X_train)\r\n```\r\n\r\n## Distance Metrics\r\n\r\nSpecialized temporal distance measures. See `references/distances.md` for complete catalog.\r\n\r\n**Usage:**\r\n```python\r\nfrom aeon.distances import dtw_distance, dtw_pairwise_distance\r\n\r\n# Single distance\r\ndistance = dtw_distance(x, y, window=0.1)\r\n\r\n# Pairwise distances\r\ndistance_matrix = dtw_pairwise_distance(X_train)\r\n\r\n# Use with classifiers\r\nfrom aeon.classification.distance_based import KNeighborsTimeSeriesClassifier\r\n\r\nclf = KNeighborsTimeSeriesClassifier(\r\n    n_neighbors=5,\r\n    distance=\"dtw\",\r\n    distance_params={\"window\": 0.2}\r\n)\r\n```\r\n\r\n**Available Distances:**\r\n- **Elastic**: DTW, DDTW, WDTW, ERP, EDR, LCSS, TWE, MSM\r\n- **Lock-step**: Euclidean, Manhattan, Minkowski\r\n- **Shape-based**: Shape DTW, SBD\r\n\r\n## Deep Learning Networks\r\n\r\nNeural architectures for time series. See `references/networks.md`.\r\n\r\n**Architectures:**\r\n- Convolutional: `FCNClassifier`, `ResNetClassifier`, `InceptionTimeClassifier`\r\n- Recurrent: `RecurrentNetwork`, `TCNNetwork`\r\n- Autoencoders: `AEFCNClusterer`, `AEResNetClusterer`\r\n\r\n**Usage:**\r\n```python\r\nfrom aeon.classification.deep_learning import InceptionTimeClassifier\r\n\r\nclf = InceptionTimeClassifier(n_epochs=100, batch_size=32)\r\nclf.fit(X_train, y_train)\r\npredictions = clf.predict(X_test)\r\n```\r\n\r\n## Datasets and Benchmarking\r\n\r\nLoad standard benchmarks and evaluate performance. See `references/datasets_benchmarking.md`.\r\n\r\n**Load Datasets:**\r\n```python\r\nfrom aeon.datasets import load_classification, load_regression\r\n\r\n# Classification\r\nX_train, y_train = load_classification(\"ArrowHead\", split=\"train\")\r\n\r\n# Regression\r\nX_train, y_train = load_regression(\"Covid3Month\", split=\"train\")\r\n```\r\n\r\n**Benchmarking:**\r\n```python\r\nfrom aeon.benchmarking import get_estimator_results\r\n\r\n# Compare with published results\r\npublished = get_estimator_results(\"ROCKET\", \"GunPoint\")\r\n```\r\n\r\n## Common Workflows\r\n\r\n### Classification Pipeline\r\n\r\n```python\r\nfrom aeon.transformations.collection import Normalizer\r\nfrom aeon.classification.convolution_based import RocketClassifier\r\nfrom sklearn.pipeline import Pipeline\r\n\r\npipeline = Pipeline([\r\n    ('normalize', Normalizer()),\r\n    ('classify', RocketClassifier())\r\n])\r\n\r\npipeline.fit(X_train, y_train)\r\naccuracy = pipeline.score(X_test, y_test)\r\n```\r\n\r\n### Feature Extraction + Traditional ML\r\n\r\n```python\r\nfrom aeon.transformations.collection import RocketTransformer\r\nfrom sklearn.ensemble import GradientBoostingClassifier\r\n\r\n# Extract features\r\nrocket = RocketTransformer()\r\nX_train_features = rocket.fit_transform(X_train)\r\nX_test_features = rocket.transform(X_test)\r\n\r\n# Train traditional ML\r\nclf = GradientBoostingClassifier()\r\nclf.fit(X_train_features, y_train)\r\npredictions = clf.predict(X_test_features)\r\n```\r\n\r\n### Anomaly Detection with Visualization\r\n\r\n```python\r\nfrom aeon.anomaly_detection import STOMP\r\nimport matplotlib.pyplot as plt\r\n\r\ndetector = STOMP(window_size=50)\r\nscores = detector.fit_predict(y)\r\n\r\nplt.figure(figsize=(15, 5))\r\nplt.subplot(2, 1, 1)\r\nplt.plot(y, label='Time Series')\r\nplt.subplot(2, 1, 2)\r\nplt.plot(scores, label='Anomaly Scores', color='red')\r\nplt.axhline(np.percentile(scores, 95), color='k', linestyle='--')\r\nplt.show()\r\n```\r\n\r\n## Best Practices\r\n\r\n### Data Preparation\r\n\r\n1. **Normalize**: Most algorithms benefit from z-normalization\r\n   ```python\r\n   from aeon.transformations.collection import Normalizer\r\n   normalizer = Normalizer()\r\n   X_train = normalizer.fit_transform(X_train)\r\n   X_test = normalizer.transform(X_test)\r\n   ```\r\n\r\n2. **Handle Missing Values**: Impute before analysis\r\n   ```python\r\n   from aeon.transformations.collection import SimpleImputer\r\n   imputer = SimpleImputer(strategy='mean')\r\n   X_train = imputer.fit_transform(X_train)\r\n   ```\r\n\r\n3. **Check Data Format**: Aeon expects shape `(n_samples, n_channels, n_timepoints)`\r\n\r\n### Model Selection\r\n\r\n1. **Start Simple**: Begin with ROCKET variants before deep learning\r\n2. **Use Validation**: Split training data for hyperparameter tuning\r\n3. **Compare Baselines**: Test against simple methods (1-NN Euclidean, Naive)\r\n4. **Consider Resources**: ROCKET for speed, deep learning if GPU available\r\n\r\n### Algorithm Selection Guide\r\n\r\n**For Fast Prototyping:**\r\n- Classification: `MiniRocketClassifier`\r\n- Regression: `MiniRocketRegressor`\r\n- Clustering: `TimeSeriesKMeans` with Euclidean\r\n\r\n**For Maximum Accuracy:**\r\n- Classification: `HIVECOTEV2`, `InceptionTimeClassifier`\r\n- Regression: `InceptionTimeRegressor`\r\n- Forecasting: `ARIMA`, `TCNForecaster`\r\n\r\n**For Interpretability:**\r\n- Classification: `ShapeletTransformClassifier`, `Catch22Classifier`\r\n- Features: `Catch22`, `TSFresh`\r\n\r\n**For Small Datasets:**\r\n- Distance-based: `KNeighborsTimeSeriesClassifier` with DTW\r\n- Avoid: Deep learning (requires large data)\r\n\r\n## Reference Documentation\r\n\r\nDetailed information available in `references/`:\r\n- `classification.md` - All classification algorithms\r\n- `regression.md` - Regression methods\r\n- `clustering.md` - Clustering algorithms\r\n- `forecasting.md` - Forecasting approaches\r\n- `anomaly_detection.md` - Anomaly detection methods\r\n- `segmentation.md` - Segmentation algorithms\r\n- `similarity_search.md` - Pattern matching and motif discovery\r\n- `transformations.md` - Feature extraction and preprocessing\r\n- `distances.md` - Time series distance metrics\r\n- `networks.md` - Deep learning architectures\r\n- `datasets_benchmarking.md` - Data loading and evaluation tools\r\n\r\n## Additional Resources\r\n\r\n- Documentation: https://www.aeon-toolkit.org/\r\n- GitHub: https://github.com/aeon-toolkit/aeon\r\n- Examples: https://www.aeon-toolkit.org/en/stable/examples.html\r\n- API Reference: https://www.aeon-toolkit.org/en/stable/api_reference.html"
  user_template: ""
variables: []
install:
  steps:
    - title: Download & unzip
      detail: >-
        Download the .zip package from SkillVault and unzip it. The aeon/ folder contains SKILL.md
        plus any companion scripts, references and assets.
    - title: Copy into your agent skills directory
      detail: >-
        Move the folder into your agent's skills directory — e.g. ~/.claude/skills/ (Claude Code),
        .codex/skills/, .cursor/skills/ or your project's .claude/skills/ folder — then restart or
        reload the session.
    - title: Use it
      detail: >-
        Describe the task in natural language; the agent loads the skill automatically when the
        request matches the skill description.
usage:
  steps: []
metadata:
  author: SkillVault
  source: https://skillvault.top/skills/aeon
  exported_at: "2026-11-03T07:41:10.794Z"