mstamp

mSTAMP

Citekey	YehEtAl2016Matrix
Source Code	stumpy
Learning type	unsupervised
Input dimensionality	multivariate

This approach uses the multidimensional matrix profile (mSTAMP). It generates an MP for every dimension and sums them up.

Output Format

The output will be an anomaly score for every input data point

Dependencies

• python 3
• numpy
• pandas
• stumpy

Notes

from timeeval.utils.window import ReverseWindowing
# post-processing for left_stampi
def post_mstamp(scores: np.ndarray, args: dict) -> np.ndarray:
    window_size = args.get("hyper_params", {}).get("anomaly_window_size", 50)
    return ReverseWindowing(window_size=window_size).fit_transform(scores)