A representation learning approach for multi-source domain adaptation achieves identifiability by partitioning the label's Markov blanket into parents, children, and spouses.
Score-based causal representation learning: Linear and general transformations
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
-
A General Representation-Based Approach to Multi-Source Domain Adaptation
A representation learning approach for multi-source domain adaptation achieves identifiability by partitioning the label's Markov blanket into parents, children, and spouses.