A dual-stream neural network framework disentangles causal features from spurious correlations in healthcare multimodal data via generalized cross-entropy loss and mutual information minimization, yielding consistent gains on MIMIC-IV, eICU, and ADNI.
Causal Analysis We formalize the data generation process and the behavior of the model using the Structured Causal Model (SCM) [19, 20, 21]
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Robust Multimodal Representation Learning in Healthcare
A dual-stream neural network framework disentangles causal features from spurious correlations in healthcare multimodal data via generalized cross-entropy loss and mutual information minimization, yielding consistent gains on MIMIC-IV, eICU, and ADNI.