Domain Aggregation Networks for Multi-Source Domain Adaptation
By Junfeng Wen, R. Greiner, Dale Schuurmans. 2019
85 citations recorded by the source corpus.
In many real-world applications, we want to exploit multiple source datasets of similar tasks to learn a model for a different but related target dataset -- e.g., recognizing characters of a new font using a set of different fonts. While most recent research has considered ad-hoc combination rules to address this problem, we extend previous work on domain discrepancy minimization to develop a finite-sample generalization bound, and accordingly propose a theoretically justified optimization procedure. The algorithm we develop, Domain AggRegation Network (DARN), is able to effectively adjust the weight of each source domain during training to ensure relevant domains are given more importance for adaptation. We evaluate the proposed method on real-world sentiment analysis and digit recognition datasets and show that DARN can significantly outperform the state-of-the-art alternatives.
Subfields: Artificial Intelligence
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Ananda Theertha Suresh
- Mehryar Mohri
- Yishay Mansour
- Ke Wu
- Corinna Cortes
- Mehrdad Mahdavi
- Csaba Szepesvari
- Qiang Yang
Citations in the corpus
- A Theory of Multiple-Source Adaptation with Limited Target Labeled Data (2020, background): “One natural approach to tackle the MSA problem we are studying consists of using discrepancy to find , by assigning a higher weight k to a source domain k that is closer to the target distribution D0 (Wen et al., 2020; Konstantinov and Lampert, 2019).”