aeon – skill Dữ liệu & Khoa học cho agent AI

Aeon Time Series Machine Learning

Overview

Aeon is a scikit-learn compatible Python toolkit for time series machine learning ([aeon-toolkit.org](https://www.aeon-toolkit.org/)). It provides algorithms across classification, regression, clustering, forecasting, anomaly detection, segmentation, similarity search, distances, transformations, benchmarking, and visualization — with a consistent estimator API.

Version note: Reviewed against aeon 1.6.0 (Python 3.13). Small synthetic checks cover classification, regression, clustering, forecasts, preprocessing, distances, search, segmentation, matrix profiles, metrics, and local dataset I/O. Remote archive and TensorFlow training snippets are illustrative; they were not executed during this review. Reference catalogs are selected methods, not exhaustive lists. See the [1.6 release notes](https://www.aeon-toolkit.org/en/stable/changelogs/v1.6.html).

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

Requires Python 3.11-3.14. Pin the reviewed release for reproducibility:

uv pip install "aeon==1.6.0"

Install only the extras required by the chosen estimator. The broad optional set is available as:

uv pip install "aeon[all_extras]==1.6.0"

On zsh, quote the extras: uv pip install "aeon[all_extras]==1.6.0".

For the matrix-profile examples: uv pip install "aeon==1.6.0" stumpy. Range precision/recall/F-score depend on prts, whose current NumPy<2 requirement conflicts with aeon 1.6; use the runnable AUC metrics or a separately validated environment. Deep learning estimators use TensorFlow. Inspect an estimator's python_dependencies tag before installing optional packages.

Experimental modules

Upstream treats forecasting, anomaly_detection, segmentation, similarity_search, and visualisation as experimental — interfaces may change between minor releases. Prefer stable modules (classification, regression, clustering, distances, transformations) for production pipelines unless you need these tasks.

Core Capabilities

1. Time Series Classification

Categorize time series into predefined classes. See references/classification.md for selected methods.

Quick Start:

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
• Accuracy candidates to validate: 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:

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:

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 (experimental module in aeon 1.x). See references/forecasting.md for forecasters.

Quick Start:

import numpy as np
from aeon.forecasting import NaiveForecaster
from aeon.forecasting.stats import ARIMA

y_train = np.array([1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10.0])

# predict returns one value at the configured horizon, not a 5-value vector
naive = NaiveForecaster(strategy="drift", horizon=5)
naive.fit(y_train)
y_at_5 = naive.predict(y_train)  # 15.0

# ARIMA uses p/d/q (not order=); multi-step via iterative_forecast
arima = ARIMA(p=1, d=1, q=1)
y_pred = arima.iterative_forecast(y_train, prediction_horizon=5)  # fits once

5. Anomaly Detection

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

Quick Start:

from aeon.anomaly_detection.series.distance_based 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:

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:

from aeon.similarity_search.subsequence import MASS

# X_train: (n_cases, n_channels, n_timepoints); query: (n_channels, 20)
searcher = MASS(length=20, normalize=True).fit(X_train)
query = X_train[0, :, :20]
indices, distances = searcher.predict(query, k=3, X_index=(0, 0))
# indices rows are (case_index, window_start); self-match is excluded

Feature Extraction and Transformations

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

ROCKET Features:

from aeon.transformations.collection.convolution_based import Rocket

rocket = Rocket()
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:

from aeon.transformations.collection.feature_based import Catch22

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

Preprocessing:

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 selected distances.

Usage:

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; temporal convolution: TCNNetwork
• Autoencoders: AEFCNClusterer, AEResNetClusterer

Usage:

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:

from aeon.datasets import load_classification, load_gunpoint, load_regression

# Classification (generic loader or dataset-specific helper)
X_train, y_train = load_classification("GunPoint", split="train")
X_train, y_train = load_gunpoint(split="train")  # same UCR dataset

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

Benchmarking:

from aeon.benchmarking.results_loaders import get_estimator_results

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

Common Workflows

Classification Pipeline

from aeon.transformations.collection import Normalizer
from aeon.classification.convolution_based import RocketClassifier
from sklearn.pipeline imp

… (nội dung đầy đủ có trong gói tải về)