load_kdd_tsad_135¶

load_kdd_tsad_135(split: Literal['train', 'test'] = 'test') → tuple[ndarray, ndarray][source]¶

Load the KDD-TSAD 135 UCR_Anomaly_Internal_Bleeding16 univariate dataset.

Returns:

Xnp.ndarray: Univariate time series with shape (n_timepoints,).
ynp.ndarray: Binary anomaly labels with shape (n_timepoints,).

Notes

The KDD-TSAD collection contains 250 datasets from different sources with a single anomaly annotated in each dataset (the top discord). It was prepared for a competition at SIGKDD 2021 [1]. Dataset 135 is a univariate dataset from this collection that records the heart rate of a patient with internal bleeding.

Dimensionality: univariate Series length: 7501 Learning Type: semi-supervised (normal training data)

References

[1]

Keogh, E., Dutta Roy, T., Naik, U. & Agrawal, A (2021). Multi-dataset Time-Series Anomaly Detection Competition, SIGKDD 2021. https://compete.hexagon-ml.com/practice/competition/39/

Examples

>>> from aeon.datasets import load_kdd_tsad_135
>>> X, y = load_kdd_tsad_135()