HNAD

Source code of Modeling Heterogeneous Normality in Time Series Anomaly Detection

Usage

1. Install dependencies in environment.yml using conda.
2. run main.py for sample usage.
3. Data set: You can find the sample input data set in the "datasets" folder. Due to the file size limit, we upload the representative subsets of TSB-AD-U and TSB-AD-M, respectively.

Major Dependencies

Python 3.8
Troch == 1.10.0+cu113
pandas == 1.3.3
scikit-learn == 0.24.2
numpy == 1.20.1