English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Zendy Plus

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Zendy Tools

Zendy Open

A biosignal is a signal that can be continuously measured from human bodies,such as respiratory sounds, heart activity (ECG), brain waves (EEG), etc, basedon which, machine learning models have been developed with very promisingperformance for automatic disease detection and health status monitoring.However, dataset shift, i.e., data distribution of inference varies from thedistribution of the training, is not uncommon for real biosignal-basedapplications. To improve the robustness, probabilistic models with uncertaintyquantification are adapted to capture how reliable a prediction is. Yet,assessing the quality of the estimated uncertainty remains a challenge. In thiswork, we propose a framework to evaluate the capability of the estimateduncertainty in capturing different types of biosignal dataset shifts withvarious degrees. In particular, we use three classification tasks based onrespiratory sounds and electrocardiography signals to benchmark fiverepresentative uncertainty quantification methods. Extensive experiments showthat, although Ensemble and Bayesian models could provide relatively betteruncertainty estimations under dataset shifts, all tested models fail to meetthe promise in trustworthy prediction and model calibration. Our work paves theway for a comprehensive evaluation for any newly developed biosignalclassifiers.

Benchmarking Uncertainty Quantification on Biosignal Classification Tasks under Dataset Shift