Wasserstein Distance Guided Representation Learning for Domain Adaptation

This paper proposes a new unsupervised domain adaptation approach called Collaborative and Adversarial Network (CAN), which uses the domain-collaborative and domain-adversarial learning strategies for training the neural network. The domain-collaborative learning strategy aims to learn domain specific feature representation to preserve the discriminability for the target domain, while the domain adversarial learning strategy aims to learn domain invariant feature representation to reduce the domain distribution mismatch between the source and target domains. We show that these two learning strategies can be uniformly formulated as domain classifier learning with positive or negative weights on the losses. We then design a collaborative and adversarial training scheme, which automatically learns domain specific representations from lower blocks in CNNs through collaborative learning and domain invariant representations from higher blocks through adversarial learning. Moreover, to further enhance the discriminability in the target domain, we propose Self-Paced CAN (SPCAN), which progressively selects pseudo-labeled target samples for re-training the classifiers. We employ a self-paced learning strategy such that we can select pseudo-labeled target samples in an easy-to-hard fashion. Additionally, we build upon the popular two-stream approach to extend our domain adaptation approach for more challenging video action recognition task, which additionally considers the cooperation between the RGB stream and the optical flow stream. We propose the Two-stream SPCAN (TS-SPCAN) method to select and reweight the pseudo labeled target samples of one stream (RGB/Flow) based on the information from the other stream (Flow/RGB) in a cooperative way. As a result, our TS-SPCAN model is able to exchange the information between the two streams. Comprehensive experiments on different benchmark datasets, Office-31, ImageCLEF-DA and VISDA-2017 for the object recognition task, and UCF101-10 and HMDB51-10 for the video action recognition task, show our newly proposed approaches achieve the state-of-the-art performance, which clearly demonstrates the effectiveness of our proposed approaches for unsupervised domain adaptation.

Electronic nose (E-nose), a kind of instrument which combines with the gas sensor and the corresponding pattern recognition algorithm, is used to detect the type and concentration of gases. However, the sensor drift will occur in realistic application scenario of E-nose, which makes a variation of data distribution in feature space and causes a decrease in prediction accuracy. Therefore, studies on the drift compensation algorithms are receiving increasing attention in the field of the E-nose. In this paper, a novel method, namely Wasserstein Distance Learned Feature Representations (WDLFR), is put forward for drift compensation, which is based on the domain invariant feature representation learning. It regards a neural network as a domain discriminator to measure the empirical Wasserstein distance between the source domain (data without drift) and target domain (drift data). The WDLFR minimizes Wasserstein distance by optimizing the feature extractor in an adversarial manner. The Wasserstein distance for domain adaption has good gradient and generalization bound. Finally, the experiments are conducted on a real dataset of E-nose from the University of California, San Diego (UCSD). The experimental results demonstrate that the effectiveness of the proposed method outperforms all compared drift compensation methods, and the WDLFR succeeds in significantly reducing the sensor drift.

A two-layer regression network for robust and accurate domain adaptation

In recent years, deep learning methods which promote the accuracy and efficiency of fault diagnosis task without any extra requirement of artificial feature extraction have elicited the attention of researchers in the field of manufacturing industry as well as aerospace. However, the problems that data in source and target domains usually have different probability distributions because of different working conditions and there are insufficient labeled or even unlabeled data in target domain significantly deteriorate the performance and generalization of deep fault diagnosis models. To address these problems, we propose a novel Wasserstein Generative Adversarial Network with Gradient Penalty- (WGAN-GP-) based deep adversarial transfer learning (WDATL) model in this study, which exploits a domain critic to learn domain invariant feature representations by minimizing the Wasserstein distance between the source and target feature distributions through adversarial training. Moreover, an improved one-dimensional convolutional neural network- (CNN-) based feature extractor which utilizes exponential linear units (ELU) as activation functions and wide kernels is designed to automatically extract the latent features of raw time-series input data. Then, the fault model classifier trained in one working condition (source domain) with sufficient labeled samples could be generalized to diagnose data in other working conditions (target domain) with insufficient labeled samples. Experiments on two open datasets demonstrate that our proposed WDATL model outperforms most of the state-of-the-art approaches on transfer diagnosis tasks under diverse working circumstances.

Domain adaptive state representation alignment for reinforcement learning

Domain adaptation is used in applying a classifier acquired in one data domain to another data domain. A classifier obtained by supervised training with labeled data in an original source domain can also be used for classification in a target domain in which the labeled data are difficult to collect with the help of domain adaptation. The most recently proposed domain adaptation methods focus on data distribution in the feature space of a classifier and bring the data distribution of both domains closer through learning. The present work is based on an existing unsupervised domain adaptation method, in which both distributions become closer through adversarial training between a target data encoder to the feature space and a domain discriminator. We propose to use the Wasserstein distance to measure the distance between two distributions, rather than the well-known Jensen-Shannon divergence. Wasserstein distance, or earth mover's distance, measures the length of the shortest path among all possible pairs between a corresponding pair of variables in two distributions. Therefore, minimization of the distance leads to overlap of the corresponding data pair in source and target domain. Thus, the classifier trained in the source domain becomes also effective in the target domain. The proposed method using Wasserstein distance shows higher accuracies in the target domains compared with an original distance in computer experiments on semantic segmentation of map images.

Domain adaptation leverages the knowledge in one domain-the source domain-to improve learning efficiency in another domain-the target domain. Existing heterogeneous domain adaptation research is relatively well-progressed but only in situations where the target domain contains at least a few labeled instances. In contrast, heterogeneous domain adaptation with an unlabeled target domain has not been well-studied. To contribute to the research in this emerging field, this article presents: 1) an unsupervised knowledge transfer theorem that guarantees the correctness of transferring knowledge and 2) a principal angle-based metric to measure the distance between two pairs of domains: one pair comprises the original source and target domains and the other pair comprises two homogeneous representations of two domains. The theorem and the metric have been implemented in an innovative transfer model, called a Grassmann-linear monotonic maps-geodesic flow kernel (GLG), which is specifically designed for heterogeneous unsupervised domain adaptation (HeUDA). The linear monotonic maps (LMMs) meet the conditions of the theorem and are used to construct homogeneous representations of the heterogeneous domains. The metric shows the extent to which the homogeneous representations have preserved the information in the original source and target domains. By minimizing the proposed metric, the GLG model learns the homogeneous representations of heterogeneous domains and transfers knowledge through these learned representations via a geodesic flow kernel (GFK). To evaluate the model, five public data sets were reorganized into ten HeUDA tasks across three applications: cancer detection, the credit assessment, and text classification. The experiments demonstrate that the proposed model delivers superior performance over the existing baselines.

This paper considers the unsupervised domain adaptation problem, in which we want to find a good prediction function on the unlabeled target domain, by utilizing the information provided in the labeled source domain. A common approach to the domain adaptation problem is to learn a representation space where the distributional discrepancy of the source and target domains is small. Existing methods generally tend to match the marginal distributions of the two domains, while the label information in the source domain is not fully exploited. In this paper, we propose a representation learning approach for domain adaptation, which is addressed as JODAWAT. We aim to adapt the joint distributions of the feature-label pairs in the shared representation space for both domains. In particular, we minimize the Wasserstein distance between the source and target domains, while the prediction performance on the source domain is also guaranteed. The proposed approach results in a minimax adversarial training procedure that incorporates a novel split gradient penalty term. A generalization bound on the target domain is provided to reveal the efficacy of representation learning for joint distribution adaptation. We conduct extensive evaluations on JODAWAT, and test its classification accuracy on multiple synthetic and real datasets. The experimental results justify that our proposed method is able to achieve superior performance compared with various domain adaptation methods.

Subdomain adaptation plays a significant role in the field of bearing fault diagnosis. It effectively aligns the pertinent distributions across subdomains and addresses the frequent issue of lacking local category information in domain adaptation. Nonetheless, this approach overlooks the quantitative discrepancies in distribution between samples from the source and target domains, leading to the vanishing gradient issue during the training of models. To tackle this challenge, this paper proposes a bearing fault diagnosis method based on Wasserstein metric residual adversarial subdomain adaptation. The Wasserstein metric is introduced as the optimized objective function of the domain discriminator in RASAN-W. The distribution discrepancy between the source domain and target domain samples is quantitatively measured, achieving the alignment of the relevant subdomain distributions between the source domain and the target domain. Ultimately, extensive experiments conducted on two real-world datasets reveal that the diagnostic accuracy of this method is significantly enhanced when compared to various leading bearing fault diagnosis techniques.

Domain adaptation attempts to transfer the knowledge obtained from the source domain to the target domain, i.e., the domain where the testing data are. The main challenge lies in the distribution discrepancy between source and target domain. Most existing works endeavor to learn domain invariant representation usually by minimizing a distribution distance, e.g., MMD and the discriminator in the recently proposed generative adversarial network (GAN). Following the similar idea of GAN, this work proposes a novel GAN architecture with duplex adversarial discriminators (referred to as DupGAN), which can achieve domain-invariant representation and domain transformation. Specifically, our proposed network consists of three parts, an encoder, a generator and two discriminators. The encoder embeds samples from both domains into the latent representation, and the generator decodes the latent representation to both source and target domains respectively conditioned on a domain code, i.e., achieves domain transformation. The generator is pitted against duplex discriminators, one for source domain and the other for target, to ensure the reality of domain transformation, the latent representation domain invariant and the category information of it preserved as well. Our proposed work achieves the state-of-the-art performance on unsupervised domain adaptation of digit classification and object recognition.

Domain Adaptation via Bregman divergence minimization

Various domain adaptation (DA) methods have been proposed to address distribution discrepancy and knowledge transfer between the source and target domains. However, many DA models focus on matching the marginal distributions of two domains and cannot satisfy fault-diagnosed-task requirements. To enhance the ability of DA, a new DA mechanism, called deep joint distribution alignment (DJDA), is proposed to simultaneously reduce the discrepancy in marginal and conditional distributions between two domains. A new statistical metric that can align the means and covariances of two domains is designed to match the marginal distributions of the source and target domains. To align the class conditional distributions, a Gaussian mixture model is used to obtain the distribution of each category in the target domain. Then, the conditional distributions of the source domain are computed via maximum-likelihood estimation, and information entropy and Wasserstein distance are employed to reduce class conditional distribution discrepancy between the two domains. With joint distribution alignment, DJDA can achieve domain confusion to the highest degree. DJDA is applied to the fault transfer diagnosis of a wind turbine gearbox and cross-bearing with unlabeled target-domain samples. Experimental results verify that DJDA outperforms other typical DA models.

Unsupervised domain adaptation (UDA) has been widely studied with the goal of transferring knowledge from a label-rich source domain to a related but unlabeled target domain. Most UDA techniques achieve this by reducing the feature discrepancies between the two domains to learn domain-invariant feature representations. While domain-invariant feature representations can reduce the differences between the source and target domains, excessively simplifying these differences may cause the model to overlook important domain-specific features, resulting in a decline in transfer learning effectiveness. To address this issue, this paper proposes a novel Gaussian Mixture Model for graph domain adaptation (GMM). This model effectively reduces the distributional bias between the source and target domains by modeling the distribution differences on a graph structure. GMM leverages the local structural information of the graph and the clustering capability of the Gaussian mixture model to automatically learn the latent mapping relationships between the source and target domains. To the best of our knowledge, this is the first work to introduce a Gaussian mixture model into UDA. Extensive experimental results on three standard benchmarks demonstrate that the proposed GMM algorithm outperforms state-of-the-art unsupervised domain adaptation methods in terms of performance.

In the absence of the labeled samples in a domain referred to as target domain, Domain Adaptation (DA) techniques come in handy. Generally, DA techniques assume there are available source domains that share similar predictive function with the target domain. Two core challenges of DA typically arise, variance that exists between source and target domains, and the inherent source hypothesis bias. In this paper, we first propose a Stability Transfer criterion for selecting relevant source domains with low variance. With this criterion, we introduce a TARget learning Assisted by Source Classifier Adaptation (TARASCA) method to address the two core challenges that have impeded the performances of DA techniques. To verify the robustness of TARASCA, extensive experimental studies are carried out with comparison to several state-of-the-art DA methods on the real-world Sentiment and Newsgroups datasets, where various settings for the class ratios of the source and target domains are considered.

Domain adaptation (DA) offers an effective way to align feature distributions of the source domain (SD) and the target domain (TD) without requiring any target label samples. As a method of DA, representation learning effectively realizes the alignment of feature distributions in different domains by transferring domain knowledge. However, existing representation learning methods often focus on unilateral representation transfer, which potentially results in transfer bias. Additionally, most methods ignore the connection between domain alignment and discrimination during the DA process, which easily causes negative transfer. This paper proposes a dynamic weighted dual-driven domain adaptation (DW-D 3 A) model that effectively addresses the aforementioned issues through bilateral feature transfer between domains and a dynamic weighted scheme. Technically, we first propose a dual-driven domain adaptation (D 3 A) model, which employs symmetrical structures to facilitate the knowledge transfer of bilateral representations between source and target domain samples, learning the subspaces of two domains and reducing distribution discrepancies between subspaces via joint distribution-driven alignment. This process mitigates transfer bias and goes beyond previous unilateral transfer methods. Then, to alleviate strong constraints on projecting SD and TD into the same subspace in existing approaches, we apply a relaxed subspace constraint to bring the projections of SD and TD closer. Furthermore, data reconstruction is incorporated to preserve discriminant information from the original data. Lastly, we expand (D 3 A) to DW-D 3 A using a dynamic weighted scheme, which adjusts the weights assigned to domain alignment and discrimination based on their significance to inhibit negative transfer. Extensive experimentation on three datasets indicates that DW-D 3 A outperforms seven other DA methods, showing its superior performance.