Browse Teach Curate

COVID-19 Cough-Sound Classification with Deep Neural Decision Forests

A COVID-19 cough classification system combining a broad audio-feature set with recursive feature elimination, deep neural decision trees/forests, Bayesian-optimized hyperparameters, SMOTE class balancing and ROC-AUC-optimal thresholding, plus a cross-dataset study showing demographic/geographic shifts impede feature transfer across cough datasets.

Data Text

Robust COVID-19 Detection from Cough Sounds using Deep Neural Decision Tree and Forest: A

The paper builds a pipeline to classify whether a recorded cough indicates COVID-19. It extracts a broad set of audio features, selects the most informative with recursive feature elimination and cro…