Improving Clinical Predictions through Unsupervised Time Series Representation Learning
This work addresses the challenge of leveraging unlabeled medical data to improve clinical decision-making, representing an incremental advancement in domain-specific methods.
The paper tackled the problem of predicting clinically relevant outcomes from medical time series by using unsupervised representation learning, showing that a forecasting Seq2Seq model with an integrated attention mechanism achieved clear performance benefits over end-to-end supervised architectures.
In this work, we investigate unsupervised representation learning on medical time series, which bears the promise of leveraging copious amounts of existing unlabeled data in order to eventually assist clinical decision making. By evaluating on the prediction of clinically relevant outcomes, we show that in a practical setting, unsupervised representation learning can offer clear performance benefits over end-to-end supervised architectures. We experiment with using sequence-to-sequence (Seq2Seq) models in two different ways, as an autoencoder and as a forecaster, and show that the best performance is achieved by a forecasting Seq2Seq model with an integrated attention mechanism, proposed here for the first time in the setting of unsupervised learning for medical time series.