We study the issue of PAC-Bayesian domain adaptation: We want to learn, from a source domain, a majority vote model dedicated to a target one. Our theoretical contribution brings a new perspective by deriving an upper-bound on the target risk where the distributions' divergence---expressed as a ratio---controls the trade-off between a source error measure and the target voters' disagreement. Our bound suggests that one has to focus on regions where the source data is informative.From this result, we derive a PAC-Bayesian generalization bound, and specialize it to linear classifiers. Then, we infer a learning algorithmand perform experiments on real data.
@article{arxiv.1506.04573,
title = {A New PAC-Bayesian Perspective on Domain Adaptation},
author = {Pascal Germain and Amaury Habrard and François Laviolette and Emilie Morvant},
journal= {arXiv preprint arXiv:1506.04573},
year = {2016}
}