AnomalyFilter: Selective Diffusion Model for Time Series Anomaly Detection

Implementation of AnomalyFilter.

Requirements

We run all the experiments in python 3.10.14, see requirements.txt for the list of pip dependencies.

To install packages

pip install -r requirements.txt

or

pip install torch==1.11.0+cu113 torchvision==0.12.0+cu113 torchaudio==0.11.0 --extra-index-url https://download.pytorch.org/whl/cu113
pip install numpy
pip install -U scikit-learn
pip install matplotlib
pip install tqdm
pip install pandas
pip install linear-attention-transformer
pip install patool

Datasets preparation

UCR Anomaly Archive

Follow the instructions on link,

and place them in ./dataset/AnomalyArchive/*.txt (e.g., /dataset/AnomalyArchive/001_UCR_Anomaly_DISTORTED1sddb40_35000_52000_52620.txt）

Or just run the code below

python main.py --dataset anomaly_archive

UCR Link: https://wu.renjie.im/research/anomaly-benchmarks-are-flawed/ and https://www.cs.ucr.edu/~eamonn/time_series_data_2018/UCR_TimeSeriesAnomalyDatasets2021.zip

---

AIOps, Yahoo

Follow the instructions on link,

Download TSB-UAD-Public.zip from TSB-UAD

and place them in ./dataset/IOPS/* or ./dataset/YAHOO/* (e.g., /dataset/IOPS/KPI-0efb375b-b902-3661-ab23-9a0bb799f4e3.test.out） (e.g., /dataset/YAHOO/Yahoo_A1real_1_data.out）

AIOps Link: https://github.com/NetManAIOps/KPI-Anomaly-Detection Yahoo Link: https://webscope.sandbox.yahoo.com/catalog.php?datatype=s&did=70

---

Server Machine Dataset (SMD)

Follow the instructions on link,

and place them in ./dataset/ServerMachineDataset/*/machine-*.txt

Or just run the code below

python main.py --dataset smd

Experiments

After the preparation of datasets, run below.

sh scripts/experiments.sh