diff --git a/docs/api_reference/anomaly_detection.rst b/docs/api_reference/anomaly_detection.rst index 92084f4c72..7db535a9be 100644 --- a/docs/api_reference/anomaly_detection.rst +++ b/docs/api_reference/anomaly_detection.rst @@ -3,58 +3,11 @@ Anomaly Detection ================= -Time Series Anomaly Detection aims at discovering regions of a time series that in -some way are not representative of the underlying generative process. -The :mod:`aeon.anomaly_detection` module contains algorithms and tools -for time series anomaly detection. The detectors have different capabilities that can -be grouped into the following categories, where ``m`` is the number of time points and -``d`` is the number of channels for a time series: +The :mod:`aeon.anomaly_detection` module contains algorithms and composition tools for time series classification. -Input data format (one of the following): - Univariate series (default): - Example: :class:`~aeon.anomaly_detection.MERLIN`. - - - np.ndarray, shape ``(m,)``, ``(m, 1)`` or ``(1, m)`` depending on axis. - - pd.DataFrame, shape ``(m, 1)`` or ``(1, m)`` depending on axis. - - pd.Series, shape ``(m,)``. - Multivariate series: - Example: :class:`~aeon.anomaly_detection.KMeansAD`. - - - np.ndarray array, shape ``(m, d)`` or ``(d, m)`` depending on axis. - - pd.DataFrame ``(m, d)`` or ``(d, m)`` depending on axis. - -Output data format (one of the following): - Anomaly scores (default): - np.ndarray, shape ``(m,)`` of type float. For each point of the input time - series, the anomaly score is a float value indicating the degree of - anomalousness. The higher the score, the more anomalous the point. The - detectors return raw anomaly scores that are not normalized. - Example: :class:`~aeon.anomaly_detection.PyODAdapter`. - Binary classification: - np.ndarray, shape ``(m,)`` of type bool or int. For each point of the input - time series, the output is a boolean or integer value indicating whether the - point is anomalous (``True``/``1``) or not (``False``/``0``). - Example: :class:`~aeon.anomaly_detection.STRAY`. - -Detector learning types: - Unsupervised (default): - Unsupervised detectors do not require any training data and can directly be - used on the target time series. You would usually call the ``fit_predict`` - method on these detectors. - Example: :class:`~aeon.anomaly_detection.DWT_MLEAD`. - Semi-supervised: - Semi-supervised detectors require a training step on a time series without - anomalies (normal behaving time series). The target value ``y`` would - consist of only zeros. You would usually first call the ``fit`` method on the - training time series and then the ``predict`` method on your target time series. - Example: :class:`~aeon.anomaly_detection.KMeansAD`. - Supervised: - Supervised detectors require a training step on a time series with known - anomalies (anomalies should be present and must be annotated). The detector - implements the ``fit`` method, and the target value ``y`` consists of zeros - and ones; ones indicating points of an anomaly. You would usually first call - the ``fit`` method on the training data and then the ``predict`` method on your - target time series. +All detectors in `aeon` can be listed using the `aeon.utils.discovery.all_estimators` utility, +using ``estimator_types="anomaly-detector"``, optionally filtered by tags. +Valid tags can be listed by calling the function `aeon.utils.discovery.all_tags_for_estimator`. Each detector in this module specifies its supported input data format, output data format, and learning type as an overview table in its documentation. Some detectors diff --git a/examples/anomaly_detection/anomaly_detection.ipynb b/examples/anomaly_detection/anomaly_detection.ipynb index a49aaa71d4..9f393437e9 100644 --- a/examples/anomaly_detection/anomaly_detection.ipynb +++ b/examples/anomaly_detection/anomaly_detection.ipynb @@ -1,35 +1,262 @@ { "cells": [ { + "cell_type": "markdown", + "id": "36c7dcfec1245abe", "metadata": {}, + "source": [ + "# Time Series Anomaly Detection with aeon\n", + "\n", + "The aim of Time Series Anomaly Detection is to discover regions of a time series that, in some way, are not representative of the underlying generative process. An anomaly in a time series can be a single point or a subsequence that deviates from the regular patterns of the sequence with respect to some measure, model or embedding. This notebook gives an overview of the anomaly detection module and the available detectors.[1]\n", + "\n", + "\"time" + ] + }, + { "cell_type": "markdown", + "id": "9673b594-aad1-46f3-bba9-158def0578fa", + "metadata": {}, "source": [ - "# Anomaly Detection\n", + "## Data Storage and Problem types\n", + "\n", + "The anomaly detectors in the `aeon.anomaly_detection` module are designed with diverse capabilities. They are categorized based on their input data format, output format, and learning type.\n", + "\n", + "### Input data format\n", + "The anomaly detectors in aeon accept time series input in either `np.ndarray` or `pd.DataFrame` formats. The shape of the input can vary depending on the number of time points (`m`) and the number of channels (`d`).\n", + "\n", + "**Univariate series (default):**\n", + "* Numpy Array: shape `(m,)`, `(m, 1)` or `(1, m)` depending on the axis.\n", + "* Pandas DataFrame: shape `(m, 1)` or `(1, m)` depending on the axis.\n", + "* Pandas Series: shape `(m,)`.\n", + "\n", + "Example: `MERLIN`.\n", + "\n", + "**Multivariate series:**\n", + "\n", + "* Numpy Array: shape `(m, d)` or `(d, m)` depending on axis.\n", + "* Pandas DataFrame: shape `(m, d)` or `(d, m)` depending on axis.\n", + "\n", + "Example: `KMeansAD`.\n", + "\n", + "### Output format\n", + "The anomaly detectors would return one of the following as output:\n", + "\n", + "**Anomaly scores (default):**\n", + "\n", + "np.ndarray, shape (m,) of type float. For each point of the input time series, the anomaly score is a float value indicating the degree of anomalousness. The higher the score, the more anomalous the point. The detectors return raw anomaly scores that are not normalized. \n", + "\n", + "Example: `PyODAdapter`.\n", + "\n", + "**Binary classification:**\n", + "\n", + "np.ndarray, shape (m,) of type bool or int. For each point of the input time series, the output is a boolean or integer value indicating whether the point is anomalous (`True`/`1`) or not (`False`/`0`). \n", + "\n", + "Example: `STRAY`.\n", "\n", - "This notebook is currently under construction! Please check back later.\n", + "Each detector in the module specifies its supported input data format, output format, and learning type as an overview table in its documentation. Some detectors support multiple learning types as well.\n", "\n", - "If you have any questions about the `aeon` anomaly detection module in the mean time, please ask us on Slack!" + "There are a couple of functions in the `aeon.datasets` module to load anomaly detection datasets. Here, we'll load the KDD-TSAD 135 UCR_Anomaly_Internal_Bleeding16 univariate dataset." + ] + }, + { + "cell_type": "code", + "execution_count": 1, + "id": "6583bc79-e76f-40a5-b2af-b4f9a4a90f74", + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } ], - "id": "36c7dcfec1245abe" + "source": [ + "import warnings\n", + "\n", + "import matplotlib.pyplot as plt\n", + "import numpy as np\n", + "\n", + "from aeon.datasets import load_kdd_tsad_135\n", + "\n", + "warnings.filterwarnings(\"ignore\")\n", + "\n", + "X, y = load_kdd_tsad_135()\n", + "\n", + "# Create a time axis\n", + "time = np.arange(len(X))\n", + "\n", + "# Separate normal and anomaly points\n", + "normal_idx = y == 0\n", + "anomaly_idx = y == 1\n", + "\n", + "# Plot the time series\n", + "plt.figure(figsize=(12, 6))\n", + "plt.plot(time, X, label=\"Time Series\", color=\"blue\", linewidth=1.5)\n", + "\n", + "# Highlight anomalies\n", + "plt.scatter(time[anomaly_idx], np.array(X)[anomaly_idx], color=\"red\", label=\"Anomalies\")\n", + "\n", + "plt.title(\"Univariate Time Series with Anomalies\")\n", + "plt.xlabel(\"Time\")\n", + "plt.ylabel(\"Value\")\n", + "plt.legend()\n", + "plt.show()" + ] + }, + { + "cell_type": "markdown", + "id": "d6701c2c-fcbd-46f0-90dd-3077337123a3", + "metadata": {}, + "source": [ + "## Anomaly Detection in aeon\n", + "\n", + "All the anomaly detectors inherit from the `BaseAnomalyDetector` class and can be categorized into one of the three categories: \n", + "\n", + "**Unsupervised (default):**\n", + "Unsupervised detectors do not require training data and can be used directly on the target time series. You would usually call the `fit_predict` method on these detectors. \n", + "\n", + "Example: `DWT_MLEAD`.\n", + "\n", + "**Semi-supervised:**\n", + "Semi-supervised detectors require a training step on a time series without anomalies (normal behaving time series). The target value `y` would consist of only zeros. You would usually first call the `fit` method on the training time series and then the `predict` method on your target time series. \n", + "\n", + "Example: `KMeansAD`.\n", + "\n", + "**Supervised:**\n", + "Supervised detectors require a training step on a time series with known anomalies (anomalies should be present and must be annotated). The detector implements the `fit` method, and the target value y consists of zeros and ones, ones indicating points of an anomaly. You would usually first call the `fit` method on the training data and then the `predict` method on your target time series.\n", + "\n", + "We currently don't have any supervised detectors. Still, the problem can be treated as an imbalanced binary classification problem, and time series classifiers can be used from the `aeon.classification` module. \n", + "\n", + "Following is the list of all the anomaly detectors available in aeon." + ] + }, + { + "cell_type": "code", + "execution_count": 2, + "id": "88276e2b-149a-4d32-aa4e-e15400c1086b", + "metadata": {}, + "outputs": [ + { + "data": { + "text/plain": [ + "[('CBLOF', aeon.anomaly_detection._cblof.CBLOF),\n", + " ('COPOD', aeon.anomaly_detection._copod.COPOD),\n", + " ('DWT_MLEAD', aeon.anomaly_detection._dwt_mlead.DWT_MLEAD),\n", + " ('IsolationForest', aeon.anomaly_detection._iforest.IsolationForest),\n", + " ('KMeansAD', aeon.anomaly_detection._kmeans.KMeansAD),\n", + " ('LOF', aeon.anomaly_detection._lof.LOF),\n", + " ('LeftSTAMPi', aeon.anomaly_detection._left_stampi.LeftSTAMPi),\n", + " ('MERLIN', aeon.anomaly_detection._merlin.MERLIN),\n", + " ('OneClassSVM', aeon.anomaly_detection._one_class_svm.OneClassSVM),\n", + " ('PyODAdapter', aeon.anomaly_detection._pyodadapter.PyODAdapter),\n", + " ('STOMP', aeon.anomaly_detection._stomp.STOMP),\n", + " ('STRAY', aeon.anomaly_detection._stray.STRAY)]" + ] + }, + "execution_count": 2, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "from aeon.utils.discovery import all_estimators\n", + "\n", + "detectors = all_estimators(\"anomaly-detector\")\n", + "detectors" + ] + }, + { + "cell_type": "markdown", + "id": "d2ce2a07-915e-4d20-94cb-0fc8aac96a8e", + "metadata": {}, + "source": [ + "For example, we have STOMP, which computes the matrix profile and records the distance of each subsequence (of a specific size) to its nearest non-self neighbour. The matrix profile can directly be interpreted as an anomaly score because a considerable distance to the closest neighbour might indicate an anomalous subsequence. We have various performance metrics such as `range_roc_auc_score` to assess the detectors." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "efd5f400-41e8-40bb-8a8e-6317386c7532", + "metadata": {}, + "outputs": [], + "source": [ + "from aeon.anomaly_detection import STOMP\n", + "from aeon.benchmarking.metrics.anomaly_detection import range_roc_auc_score\n", + "\n", + "detector = STOMP(window_size=200)\n", + "scores = detector.fit_predict(X)\n", + "y_pred = detector.fit_predict(X)\n", + "range_roc_auc_score(y, y_pred)" + ] + }, + { + "cell_type": "markdown", + "id": "bcfcbfc5-8d6e-4a5d-ba3c-b952c32dc8f3", + "metadata": {}, + "source": [ + "Another example is the `PyODAdapter`, which allows us to use all outlier detection methods of [PyOD](https://pyod.readthedocs.io/en/latest/) for time series anomaly detection." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "743fbbaa-a7d0-4f56-993a-07453f6a9442", + "metadata": {}, + "outputs": [], + "source": [ + "from pyod.models.ocsvm import OCSVM\n", + "\n", + "from aeon.anomaly_detection import PyODAdapter\n", + "from aeon.benchmarking.metrics.anomaly_detection import range_roc_auc_score\n", + "\n", + "detector = PyODAdapter(OCSVM(), window_size=3)\n", + "y_scores = detector.fit_predict(X, axis=0)\n", + "range_roc_auc_score(y, y_scores)" + ] + }, + { + "cell_type": "markdown", + "id": "bb927d1a-ded3-4b94-8c03-604eb40f104c", + "metadata": {}, + "source": [ + "## References\n", + "\n", + "[1] Sebastian Schmidl, Phillip Wenig, and Thorsten Papenbrock. Anomaly\n", + " Detection in Time Series: A Comprehensive Evaluation. PVLDB, 15(9): 1779- 1797, 2022.\n", + " doi:10.14778/3538598.3538602" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "6083369f-fb43-42de-a2c2-58308a3977cf", + "metadata": {}, + "outputs": [], + "source": [] } ], "metadata": { "kernelspec": { - "display_name": "Python 3", + "display_name": "Python 3 (ipykernel)", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", - "version": 2 + "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", - "pygments_lexer": "ipython2", - "version": "2.7.6" + "pygments_lexer": "ipython3", + "version": "3.11.10" } }, "nbformat": 4, diff --git a/examples/anomaly_detection/img/anomaly_detection.png b/examples/anomaly_detection/img/anomaly_detection.png index 3e8d78bcfa..9e950da16f 100644 Binary files a/examples/anomaly_detection/img/anomaly_detection.png and b/examples/anomaly_detection/img/anomaly_detection.png differ