A Bayesian method that jointly models multiple outcomes and sequentially updates across datasets with a learned weight that limits negative transfer.
Unlike univariate RECaST, the multivariate random effect βi does not always follow a multivariate Cauchy distribution, even in the canonical case
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ME 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Multivariate and Online Transfer Learning with Uncertainty Quantification
A Bayesian method that jointly models multiple outcomes and sequentially updates across datasets with a learned weight that limits negative transfer.