On the Benefits of Selectivity in Pseudo-Labeling for Unsupervised Multi-Source-Free Domain Adaptation.

Maohao Shen Yuheng Bu Gregory W. Wornell

Published in: CoRR (2022)

Keyphrases

domain adaptation
multi source
multiple sources
semi supervised
manifold alignment
unsupervised learning

data fusion
target domain
information fusion
data sources
active learning
semi supervised learning

labeled data
data integration
cross domain
co training
user interests
sentiment classification

supervised learning
community detection
unlabeled data
social networks
document classification