This repository contains the source code for our paper Unsupervised Domain Adaptation for Cross-Patient Seizure Classification (JNE, 2023).
📰 News: The original EEG trials of our seizure dataset CHSZ are now publicly available on Zenodo.
- tasa_sds_gda merges the proposed tasa, sds, and gda methods;
- mldg and maml are two meta-learning baselines;
- mlp is a deep neural network baseline without using any tricks.
Dataset can be obtained in directory "./data/fts_labels/", containing S1-S27.
For more details regarding the original EEG signals from the CHSZ dataset, please contact via email at [email protected].
Some necessary functions are in utils directory.
If you find this repo helpful, please cite our work:
@article{Wang2023TASA,
title={Unsupervised domain adaptation for cross-patient seizure classification},
author={Wang, Ziwei and Zhang, Wen and Li, Siyang and Chen, Xinru and Wu, Dongrui},
journal={Journal of Neural Engineering},
volume={20},
number={6},
pages={066002},
year={2023},
}