Robust unsupervised domain adaptation for neural networks via moment alignment

Autoren Werner Zellinger
Bernhard A. Moser
Thomas Grubinger
Edwin Lughofer
Thomas Natschläger
Susanne Saminger-Platz
Editoren
Titel Robust unsupervised domain adaptation for neural networks via moment alignment
Typ Artikel
Journal Information Sciences
Verlag Elsevier
Nummer 5
Band 483
DOI 10.1016/j.ins.2019.01.025
Monat May
Jahr 2019
Seiten 174-191
SCCH ID# 17094
Abstract

A novel approach for unsupervised domain adaptation for neural networks is proposed. It relies on metric-based regularization of the learning process. The metric-based regularization aims at domain-invariant latent feature representations by means of maximizing the similarity between domain-specific activation distributions. The proposed metric results from modifying an integral probability metric such that it becomes less translation-sensitive on a polynomial function space. The metric has an intuitive interpretation in the dual space as the sum of differences of higher order central moments of the corresponding activation distributions. Under appropriate assumptions on the input distributions, error minimization is proven for the continuous case. As demonstrated by an analysis of standard benchmark experiments for sentiment analysis, object recognition and digit recognition, the outlined approach is robust regarding parameter changes and achieves higher classification accuracies than comparable approaches.