PAC-Bayesian Theorems for Domain Adaptation with Specialization to Linear Classifiers
Abstract
In this paper, we provide two main contributions in PAC-Bayesian theory for domain adaptation where the objective is to learn, from a source distribution, a well-performing majority vote on a different target distribution. On the one hand, we propose an improvement of the previous approach proposed by Germain et al. (2013), that relies on a novel distribution pseudodistance based on a disagreement averaging, allowing us to derive a new tighter PAC-Bayesian domain adaptation bound for the stochastic Gibbs classifier. We specialize it to linear classifiers, and design a learning algorithm which shows interesting results on a synthetic problem and on a popular sentiment annotation task. On the other hand, we generalize these results to multisource domain adaptation allowing us to take into account different source domains. This study opens the door to tackle domain adaptation tasks by making use of all the PAC-Bayesian tools.
Cite
@article{arxiv.1503.06944,
title = {PAC-Bayesian Theorems for Domain Adaptation with Specialization to Linear Classifiers},
author = {Pascal Germain and Amaury Habrard and François Laviolette and Emilie Morvant},
journal= {arXiv preprint arXiv:1503.06944},
year = {2016}
}
Comments
This report is a long version of our paper entitled A PAC-Bayesian Approach for Domain Adaptation with Specialization to Linear Classifiers published in the proceedings of the International Conference on Machine Learning (ICML) 2013. We improved our main results, extended our experiments, and proposed an extension to multisource domain adaptation