This is the accompanying repository of our paper "Shapelet-based Temporal Association Rule Mining for Multivariate Time Series Classification" presented at the SIGKDD 2022 Workshop on Mining and Learning from Time Series (MiLeTS).

The packages required to run this code are listed in requirements.txt. To create a new virtual environment and install them:

python3 -m pip install --user virtualenv
python3 -m venv sets
source sets/bin/activate
pip install -r requirements.txt

The solar flare dataset is provided in the data/sf directory.

sets.sh runs SETS on the solar flare dataset as described in the paper. Feel free to experiment with different datasets and parameters.

To use a custom dataset, split it into train and test sets as 3D Numpy arrays with shape (N,D,L), such that N is the number of time series instances, D is the number of dimensions, and L is the time series length, and save it in a new directory under data.

chmod +x sets.sh
./sets.sh

For large datasets, and depending on the time contract, parts of SETS might take longer to run. sets.sh keeps intermediary results to allow reusing them if needed.