Aggregating From Multiple Target-Shifted Sources
arXiv:2105.04051
Abstract
Multi-source domain adaptation aims at leveraging the knowledge from multiple tasks for predicting a related target domain. Hence, a crucial aspect is to properly combine different sources based on their relations. In this paper, we analyzed the problem for aggregating source domains with different label distributions, where most recent source selection approaches fail. Our proposed algorithm differs from previous approaches in two key ways: the model aggregates multiple sources mainly through the similarity of semantic conditional distribution rather than marginal distribution; the model proposes a \emph{unified} framework to select relevant sources for three popular scenarios, i.e., domain adaptation with limited label on target domain, unsupervised domain adaptation and label partial unsupervised domain adaption. We evaluate the proposed method through extensive experiments. The empirical results significantly outperform the baselines.
References in corpus (21)
- An Overview of Multi-Task Learning in Deep Neural Networks
- A Survey on Multi-Task Learning
- Marginalized Denoising Autoencoders for Domain Adaptation
- Aligning Domain-specific Distribution and Classifier for Cross-domain Classification from Multiple Sources
- Multi-source Transfer Learning with Convolutional Neural Networks for Lung Pattern Analysis
- Parameter-Efficient Transfer Learning for NLP
- A Convex Formulation for Learning Task Relationships in Multi-Task Learning
- Algorithms and Theory for Multiple-Source Adaptation
- Multi-Adversarial Domain Adaptation
- DIVA: Domain Invariant Variational Autoencoders
- Multiple Source Adaptation and the Renyi Divergence
- Multi-source Domain Adaptation in the Deep Learning Era: A Systematic Survey
- Domain Adaptation with Conditional Distribution Matching and Generalized Label Shift
- Regularized Learning for Domain Adaptation under Label Shifts
- Robust Learning from Untrusted Sources
- A Unified View of Label Shift Estimation
- Multi-source Distilling Domain Adaptation
- On Target Shift in Adversarial Domain Adaptation
- Normalized Wasserstein Distance for Mixture Distributions with Applications in Adversarial Learning and Domain Adaptation
- A Theory of Multiple-Source Adaptation with Limited Target Labeled Data
- Multi-source Domain Adaptation for Visual Sentiment Classification