Matches in SemOpenAlex for { <https://semopenalex.org/work/W4387687079> ?p ?o ?g. }
Showing items 1 to 67 of
67
with 100 items per page.
- W4387687079 abstract "Classification models for electroencephalogram (EEG) data show a large decrease in performance when evaluated on unseen test sub jects. We reduce this performance decrease using new regularization techniques during model training. We propose several graphical models to describe an EEG classification task. From each model, we identify statistical relationships that should hold true in an idealized training scenario (with infinite data and a globally-optimal model) but that may not hold in practice. We design regularization penalties to enforce these relationships in two stages. First, we identify suitable proxy quantities (divergences such as Mutual Information and Wasserstein-1) that can be used to measure statistical independence and dependence relationships. Second, we provide algorithms to efficiently estimate these quantities during training using secondary neural network models. We conduct extensive computational experiments using a large benchmark EEG dataset, comparing our proposed techniques with a baseline method that uses an adversarial classifier. We find our proposed methods significantly increase balanced accuracy on test subjects and decrease overfitting. The proposed methods exhibit a larger benefit over a greater range of hyperparameters than the baseline method, with only a small computational cost at training time. These benefits are largest when used for a fixed training period, though there is still a significant benefit for a subset of hyperparameters when our techniques are used in conjunction with early stopping regularization." @default.
- W4387687079 created "2023-10-17" @default.
- W4387687079 creator A5023338067 @default.
- W4387687079 creator A5024747748 @default.
- W4387687079 creator A5033810782 @default.
- W4387687079 creator A5034909011 @default.
- W4387687079 creator A5048279362 @default.
- W4387687079 creator A5065592637 @default.
- W4387687079 creator A5083261801 @default.
- W4387687079 date "2023-10-12" @default.
- W4387687079 modified "2023-10-18" @default.
- W4387687079 title "Stabilizing Subject Transfer in EEG Classification with Divergence Estimation" @default.
- W4387687079 doi "https://doi.org/10.48550/arxiv.2310.08762" @default.
- W4387687079 hasPublicationYear "2023" @default.
- W4387687079 type Work @default.
- W4387687079 citedByCount "0" @default.
- W4387687079 crossrefType "posted-content" @default.
- W4387687079 hasAuthorship W4387687079A5023338067 @default.
- W4387687079 hasAuthorship W4387687079A5024747748 @default.
- W4387687079 hasAuthorship W4387687079A5033810782 @default.
- W4387687079 hasAuthorship W4387687079A5034909011 @default.
- W4387687079 hasAuthorship W4387687079A5048279362 @default.
- W4387687079 hasAuthorship W4387687079A5065592637 @default.
- W4387687079 hasAuthorship W4387687079A5083261801 @default.
- W4387687079 hasBestOaLocation W43876870791 @default.
- W4387687079 hasConcept C118552586 @default.
- W4387687079 hasConcept C119857082 @default.
- W4387687079 hasConcept C153180895 @default.
- W4387687079 hasConcept C154945302 @default.
- W4387687079 hasConcept C15744967 @default.
- W4387687079 hasConcept C22019652 @default.
- W4387687079 hasConcept C2776135515 @default.
- W4387687079 hasConcept C41008148 @default.
- W4387687079 hasConcept C50644808 @default.
- W4387687079 hasConcept C522805319 @default.
- W4387687079 hasConcept C5465570 @default.
- W4387687079 hasConcept C8642999 @default.
- W4387687079 hasConcept C95623464 @default.
- W4387687079 hasConceptScore W4387687079C118552586 @default.
- W4387687079 hasConceptScore W4387687079C119857082 @default.
- W4387687079 hasConceptScore W4387687079C153180895 @default.
- W4387687079 hasConceptScore W4387687079C154945302 @default.
- W4387687079 hasConceptScore W4387687079C15744967 @default.
- W4387687079 hasConceptScore W4387687079C22019652 @default.
- W4387687079 hasConceptScore W4387687079C2776135515 @default.
- W4387687079 hasConceptScore W4387687079C41008148 @default.
- W4387687079 hasConceptScore W4387687079C50644808 @default.
- W4387687079 hasConceptScore W4387687079C522805319 @default.
- W4387687079 hasConceptScore W4387687079C5465570 @default.
- W4387687079 hasConceptScore W4387687079C8642999 @default.
- W4387687079 hasConceptScore W4387687079C95623464 @default.
- W4387687079 hasLocation W43876870791 @default.
- W4387687079 hasOpenAccess W4387687079 @default.
- W4387687079 hasPrimaryLocation W43876870791 @default.
- W4387687079 hasRelatedWork W1724442104 @default.
- W4387687079 hasRelatedWork W2156876426 @default.
- W4387687079 hasRelatedWork W2922073769 @default.
- W4387687079 hasRelatedWork W3008105014 @default.
- W4387687079 hasRelatedWork W3155135229 @default.
- W4387687079 hasRelatedWork W4288022214 @default.
- W4387687079 hasRelatedWork W4298369531 @default.
- W4387687079 hasRelatedWork W4320854463 @default.
- W4387687079 hasRelatedWork W4385768004 @default.
- W4387687079 hasRelatedWork W3040157805 @default.
- W4387687079 isParatext "false" @default.
- W4387687079 isRetracted "false" @default.
- W4387687079 workType "article" @default.