A Survey on Adversarial Domain Adaptation

Abstract

Having a lot of labeled data is always a problem in machine learning issues. Even by collecting lots of data hardly, shift in data distribution might emerge because of differences in source and target domains. The shift would make the model to face with problems in test step. Therefore, the necessity of using domain adaptation emerges. There are three techniques in the field of domain adaptation namely discrepancy based, adversarial based and reconstruction based methods. For domain adaptation, adversarial learning approaches showed state-of-the-art performance. Although there are some comprehensive surveys about domain adaptation, we technically focus on adversarial based domain adaptation methods. We examine each proposed method in detail with respect to their structures and objective functions. The common aspect of proposed methods besides domain adaptation is considering the target labels are predicted as accurately as possible. It can be represented by some methods such as metric learning and multi-adversarial discriminators as are used in some of the papers. Also, we address the negative transfer issue for dissimilar distributions and propose the addition of clustering heuristics to the underlying structures for future research.

Multi-Perception Graph Convolution Transfer Network Bearing Fault Diagnosis Method

“…Unsupervised domain adaptation is an important research field in domain adaptation [22]. In order to illustrate this problem clearly, the basic standard notation of domain adaptation is introduced.…”

Triple Loss Adversarial Domain Adaptation Network for Cross-Domain Sea–Land Clutter Classification

“…DA-based methods have made significant progress in fields related to sea-land clutter or OTHR, such as computer vision (CV), fault diagnosis, and remote sensing. In general, according to different strategies for aligning feature distributions, DA-based methods can be divided into three categories [13], [14]: 1) discrepancy-based DA; 2) reconstruction-based DA; and 3) adversarial-based DA. The discrepancy-based DA aims to reduce the distance or divergence between different distributions by selecting certain metrics, such as maximum mean discrepancy (MMD) [15], multiple kernel MMD (MK-MMD) [16], deep correlation alignment (Deep CORAL) [17], and central moment discrepancy (CMD) [18].…”