GNN-based Time Series Segmentation

This project implements a graph-based approach to time series segmentation (TSS) using Graph Neural Networks (GNNs). It leverages synthetic and real-world datasets from the Time Series Segmentation Benchmark (TSSB) (https://github.com/ermshaua/time-series-segmentation-benchmark/blob/main/README.md) to train a multi-layer Graph Attention Network (GAT). The repository includes both a Jupyter Notebook that demonstrates the full segmentation pipeline — from data loading and graph construction to model training and evaluation — and an environment.yml file for easy environment setup. The enviroment.yml file includes also the dataset package doirect

Envrionment Setup

1. Use mamba (conda) to build the environment from the provided environment.yml : mamba env create -f environment.yml

2. Activate the environment: conda activate TSSB

3. Install ipykernel and add the environment as a Jupyter kernel: pip install ipykernel python -m ipykernel install --user --name TSSB --display-name "Python (TSSB)"

4. Launch Jupyter Notebook and select the Python (TSSB) kernel

Citation

License

Acknowledgement

This work was supported by the Slovenian Research Agency (P2-0016) and the European Commission NANCY project (No.101096456).