Multi-view self-supervised learning for multivariate variable-channel time series
This work addresses a bottleneck in applying self-supervised learning to biomedical time series where channel sets differ, offering a solution for more flexible model transfer in this domain.
The paper tackles the problem of transferring self-supervised learning models across multivariate time series datasets with varying input channels by proposing an encoder that operates on individual channels and uses a message passing neural network to combine them. The method, when pretrained on a six-channel EEG dataset and fine-tuned on a two-channel dataset, outperforms other approaches in most settings, particularly with the TS2Vec loss.
Labeling of multivariate biomedical time series data is a laborious and expensive process. Self-supervised contrastive learning alleviates the need for large, labeled datasets through pretraining on unlabeled data. However, for multivariate time series data, the set of input channels often varies between applications, and most existing work does not allow for transfer between datasets with different sets of input channels. We propose learning one encoder to operate on all input channels individually. We then use a message passing neural network to extract a single representation across channels. We demonstrate the potential of this method by pretraining our model on a dataset with six EEG channels and then fine-tuning it on a dataset with two different EEG channels. We compare models with and without the message passing neural network across different contrastive loss functions. We show that our method, combined with the TS2Vec loss, outperforms all other methods in most settings.