Multi-source Domain Research Articles

Class-aware sample reweighting optimal transport for multi-source domain adaptation

Multi-Source Domain Adaptation (MSDA) techniques have attracted widespread attention due to their availability to transfer knowledge from multiple source domains to the unlabeled target domain. Optimal transport (OT) has recently been utilized to measure the distance between distributions in virtue of its robustness. This paper proposes a novel OT-based Class-Aware Sample Reweighting (CASR) method to achieve sample-level fine-grained alignment between multi-source and target. Technically, the class-aware sampling strategy ensures class-level conditional alignment during transport by explicitly selecting samples from each domain. Besides, the sample-reweighting module is designed to allocate specific mass to each transmitted sample, which considers the classification reliability and the spatial information correlation to obtain the alignment priority between target and multi-source and further optimize the transport plan. Extensive experiments conducted on several benchmarks show that CASR presents significant advantages compared with other MSDA methods, and the visualization analysis further demonstrates the effectiveness of each proposed module.

Riemannian representation learning for multi-source domain adaptation

Multi-Source Domain Adaptation (MSDA) aims at training a classification model that achieves small target error, by leveraging labeled data from multiple source domains and unlabeled data from a target domain. The source and target domains are described by related but different joint distributions, which lie on a Riemannian manifold named the statistical manifold. In this paper, we characterize the joint distribution difference by the Hellinger distance, which bears strong connection to the Riemannian metric defined on the statistical manifold. We show that the target error of a neural network classification model is upper bounded by the average source error of the model and the average Hellinger distance, i.e., the average of multiple Hellinger distances between the source and target joint distributions in the network representation space. Motivated by the error bound, we introduce Riemannian Representation Learning (RRL): An approach that trains the network model by minimizing (i) the average empirical Hellinger distance with respect to the representation function, and (ii) the average empirical source error with respect to the network model. Specifically, we derive the average empirical Hellinger distance by constructing and solving unconstrained convex optimization problems whose global optimal solutions are easy to find. With the network model trained, we expect it to achieve small error in the target domain. Our experimental results on several image datasets demonstrate that the proposed RRL approach is statistically better than the comparison methods.

A novel conditional weighting transfer Wasserstein auto-encoder for rolling bearing fault diagnosis with multi-source domains

Transfer learning based on a single source domain to a target domain has received a lot of attention in the cross-domain fault diagnosis tasks of rolling bearing. However, the practical issues often contain multiple source domain data, and the information contained in the target domain is quite different from one source domain to another. Therefore, the transfer pattern from multiple source domains to the target domain undoubtedly has brighter application prospects. Based on these discussions, a multi-source domain transfer learning approach called conditional weighting transfer Wasserstein auto-encoder is developed to deal with the challenges of cross-domain fault diagnosis. Different from the traditional distribution alignment idea of directly aligning the source and target domains, the proposed framework adopts an indirect latent alignment idea to achieve better feature alignment, that is, indirectly aligning the feature distribution of source and target in the latent feature space with the help of Gaussian prior distribution. Furthermore, considering the variability of different source domains containing information about the target domain, an ingenious conditional weighting strategy is designed to quantify the similarity of different source domains to target domain, and further help the proposed model to minimize the discrepancy in conditional distribution. The cross-domain fault diagnosis tasks adequately verify that the proposed framework can sufficiently transfer knowledge from all source domains to the target domain, and has extensive application prospects.

Multi-source domain adaptation of GPR data for IED detection

Multi-source domain adaptation of GPR data for IED detection

Multi-source domain adaptation method for textual emotion classification using deep and broad learning

Existing domain adaptation methods for classifying textual emotions have the propensity to focus on single-source domain exploration rather than multi-source domain adaptation. The efficacy of emotion classification is hampered by the restricted information and volume from a single source domain. Thus, to improve the performance of domain adaptation, we present a novel multi-source domain adaptation approach for emotion classification, by combining broad learning and deep learning in this article. Specifically, we first design a model to extract domain-invariant features from each source domain to the same target domain by using BERT and Bi-LSTM, which can better capture contextual features. Then we adopt broad learning to train multiple classifiers based on the domain-invariant features, which can more effectively conduct multi-label classification tasks. In addition, we design a co-training model to boost these classifiers. Finally, we carry out several experiments on four datasets by comparison with the baseline methods. The experimental results show that our proposed approach can significantly outperform the baseline methods for textual emotion classification.

Domain-specific feature elimination: multi-source domain adaptation for image classification

Domain-specific feature elimination: multi-source domain adaptation for image classification

Deep Negative Correlation Multisource Domains Adaptation Network for Machinery Fault Diagnosis Under Different Working Conditions

Machinery fault diagnosis is crucial to ensure production safety. Deep learning has been widely investigated in intelligent fault diagnosis. However, the mechanical equipment generally runs in a complex and polytropic condition. It is challenging to perform machinery fault diagnosis under polytopic conditions. In this article, a novel multisource domain adaptation method is proposed for machinery fault diagnosis under different working conditions. First, a deep negative correlation multisource domains adaptation network (DNC-MDAN) is proposed, where the information from multisource domains is transferred into target domain based on the multisource feature alignment and adversarial learning. Second, in order to make full of the multisource domains information, an ensemble classifier corresponding to multisource domains is developed, where a DNC learning is performed in the ensemble classifier. Third, an end-to-end feature generator, discrete cosine convolutional block, is proposed in DNC-MDAN, where discrete cosine transform is embedded in the network smoothly for noise reduction. In order to validate the effectiveness of DNC-MDAN for machinery fault diagnosis, three cases are considered in this article. In comparison with other transfer learning methods, DNC-MDAN has a better fault diagnosis performance.

Dual collaboration for decentralized multi-source domain adaptation

Dual collaboration for decentralized multi-source domain adaptation

A no-free-lunch theorem for multitask learning

Multitask learning and related areas such as multisource domain adaptation address modern settings where data sets from N related distributions {Pt} are to be combined toward improving performance on any single such distribution D. A perplexing fact remains in the evolving theory on the subject: while we would hope for performance bounds that account for the contribution from multiple tasks, the vast majority of analyses result in bounds that improve at best in the number n of samples per task, but most often do not improve in N. As such, it might seem at first that the distributional settings or aggregation procedures considered in such analyses might be somehow unfavorable; however, as we show, the picture happens to be more nuanced, with interestingly hard regimes that might appear otherwise favorable. In particular, we consider a seemingly favorable classification scenario where all tasks Pt share a common optimal classifier h∗, and which can be shown to admit a broad range of regimes with improved oracle rates in terms of N and n. Some of our main results are: ∙ We show that, even though such regimes admit minimax rates accounting for both n and N, no adaptive algorithm exists, that is, without access to distributional information, no algorithm can guarantee rates that improve with large N for n fixed. ∙ With a bit of additional information, namely, a ranking of tasks {Pt} according to their distance to a target D, a simple rank-based procedure can achieve near optimal aggregations of tasks’ data sets, despite a search space exponential in N. Interestingly, the optimal aggregation might exclude certain tasks, even though they all share the same h∗.

Multi-Source Domain Transfer Discriminative Dictionary Learning Modeling for Electroencephalogram-Based Emotion Recognition

Cognitive computing is dedicated to researching a computing principle and method that can simulate the intelligence ability of human brain. Human emotion is the basic component of human cognitive activities. Electroencephalogram (EEG) computer signals obtained from a brain computer interface are difficult to conceal, and using machine learning methods to analyze EEG emotion is a hot topic in artificial intelligence. However, the EEG signal is non-stationary, making it difficult to select sufficient data from the same person to train a classifier for a subject. To promote the performance of emotion recognition methods, a multi-source domain transfer discriminative dictionary learning modeling (MDTDDL) is proposed in this study. The method integrates transfer learning and dictionary learning in a learning model, including the concepts of subspace learning, manifold smoothness, margin-based discriminant embedding, and large margin. The domain-specific transformation matrix projects EEG signals from various domains into the transfer subspace. The domain-invariant dictionary can find potential connections between multiple source domains and target domain. The manifold smoothness and margin-based discriminant embedding term further improve the model’s learning ability. The alternating optimization technique is used in model solving to efficiently compute model parameters. Experiments on the SEED and DEAP datasets demonstrate the effectiveness of MDTDDL.

Unsupervised domain adaptation based on the predictive uncertainty of models

Unsupervised domain adaptation (UDA) aims to improve the prediction performance in the target domain under distribution shifts from the source domain. The key principle of UDA is to minimize the divergence between the source and the target domains. To follow this principle, many methods employ a domain discriminator to match the feature distributions. Some recent methods evaluate the discrepancy between two predictions on target samples to detect those that deviate from the source distribution. However, their performance is limited because they either match the marginal distributions or measure the divergence conservatively. In this paper, we present a novel UDA method that learns domain-invariant features that minimize the domain divergence. We propose model uncertainty as a measure of the domain divergence. Our UDA method based on model uncertainty (MUDA) adopts a Bayesian framework and provides an efficient way to evaluate model uncertainty by means of Monte Carlo dropout sampling. Experiment results on image recognition tasks show that our method is superior to existing state-of-the-art methods. We also extend MUDA to multi-source domain adaptation problems.

Parallel learned generative adversarial network with multi-path subspaces for cross-modal retrieval

Cross-modal retrieval aims to narrow the heterogeneity gap between different modalities, such as retrieving images through texts or vice versa. One of the key challenges of cross-modal retrieval is the inconsistent distribution across diverse modalities. Most existing methods tend to construct a common representation subspace to overcome the challenge. However, the supervision information is not fully explored in most single-path cross-modal learning approaches. In this paper, we present a novel Parallel Learned generative adversarial network with Multi-path Subspaces (PLMS) for cross-modal retrieval. PLMS is a parallel learned architecture that aims to capture more effective information in an end-to-end trained cross-modal retrieval model. To be specific, a dual-branch network is constructed in the modality-specific generator, thereby the overall framework learns two common subspaces to emphasize discrepant supervision information and preserve more effective transformed features. We further design two objective functions for the training of the dual branches in generators. Through joint training, the feature representations generated by dual branches in a specific modality are fused for similarity measurement between modalities. To avoid redundancy and overlap during fusion, a Multi-source Domain Balancing (MDB) mechanism is presented to explore the contribution of the two specific-task branches. Extensive experiments show that our proposed method is effective and achieves state-of-the-art results on four widely-used databases.

Transferable feature filtration network for multi-source domain adaptation

Compared to conventional unsupervised domain adaptation, multi-source domain adaptation (MSDA) is suitable for more practical scenarios but is more challenging because the knowledge transferred to the target domain is learned from multiple source domains. Existing adversarial-based domain adaptation methods facilitate the obfuscation of domain distributions and cross-domain alignment for representations by deceiving the discriminator. However, traditional methods neglect the distinction in the transferability of the different features, resulting in untransferable domain-varying features, such as those extracted from the background, also being forced to align between domains. To this end, we propose a feature filtration mechanism and design a corresponding neural network to achieve a selective feature alignment based on the transferability of features; this is termed a transferable feature filtration network (TFFN). We construct a transferable feature learning framework containing two subprocesses, filtering, in which the attention mechanism-based filtration network automatically extracts transferable features under additional supervision, and repairing, in which we suggest enhancing the adaptability of the filtration network by repairing it with the help of an additional classifier focused on the target knowledge. Furthermore, to facilitate the matching of the source and target distributions at the class level, we propose a filtration consistency loss to enhance the cross-domain consistency of the filtering weights. Extensive experiments conducted on MSDA benchmark datasets show that the proposed method has significant advantages over the existing methods.

Disk anomaly detection for remote maintenance control system of natural gas pipeline based on multi-source domain transfer and incremental learning

Accurate disk anomaly detection is of great significance for maintaining the safe and stable operation of the remote control system of natural gas pipelines. Most disk transfer algorithms use only a single-source domain model for offline transfer learning, resulting in lower classifier performance. This paper proposes a disk online anomaly detection algorithm combining transfer learning and incremental learning. A transfer learning algorithm based on data distribution similarity and dynamic index weighted ensemble is proposed. According to multiple source domain models and target domain model negative data in hyperspace, k neighbour data of each target domain negative data is extracted, and then different test criteria are adopted for each sub-model according to the number of target domain positive data, and index reweighting is performed according to the model detection capability. An incremental learning algorithm based on online automatic annotation is proposed. The data popped up by the sliding window is automatically marked at the current disk state, and the update period of the prediction model is dynamically adjusted according to the false detection rate and the missed detection rate. By setting three different experiments for comparison, the effectiveness and advanced nature of the method proposed in this paper are illustrated.

Multi-Source Contribution Learning for Domain Adaptation.

Transfer learning becomes an attractive technology to tackle a task from a target domain by leveraging previously acquired knowledge from a similar domain (source domain). Many existing transfer learning methods focus on learning one discriminator with single-source domain. Sometimes, knowledge from single-source domain might not be enough for predicting the target task. Thus, multiple source domains carrying richer transferable information are considered to complete the target task. Although there are some previous studies dealing with multi-source domain adaptation, these methods commonly combine source predictions by averaging source performances. Different source domains contain different transferable information; they may contribute differently to a target domain compared with each other. Hence, the source contribution should be taken into account when predicting a target task. In this article, we propose a novel multi-source contribution learning method for domain adaptation (MSCLDA). As proposed, the similarities and diversities of domains are learned simultaneously by extracting multi-view features. One view represents common features (similarities) among all domains. Other views represent different characteristics (diversities) in a target domain; each characteristic is expressed by features extracted in a source domain. Then multi-level distribution matching is employed to improve the transferability of latent features, aiming to reduce misclassification of boundary samples by maximizing discrepancy between different classes and minimizing discrepancy between the same classes. Concurrently, when completing a target task by combining source predictions, instead of averaging source predictions or weighting sources using normalized similarities, the original weights learned by normalizing similarities between source and target domains are adjusted using pseudo target labels to increase the disparities of weight values, which is desired to improve the performance of the final target predictor if the predictions of sources exist significant difference. Experiments on real-world visual data sets demonstrate the superiorities of our proposed method.

Prototype-Based Multisource Domain Adaptation.

Unsupervised domain adaptation aims to transfer knowledge from labeled source domain to unlabeled target domain. Recently, multisource domain adaptation (MDA) has begun to attract attention. Its performance should go beyond simply mixing all source domains together for knowledge transfer. In this article, we propose a novel prototype-based method for MDA. Specifically, for solving the problem that the target domain has no label, we use the prototype to transfer the semantic category information from source domains to target domain. First, a feature extraction network is applied to both source and target domains to obtain the extracted features from which the domain-invariant features and domain-specific features will be disentangled. Then, based on these two kinds of features, the named inherent class prototypes and domain prototypes are estimated, respectively. Then a prototype mapping to the extracted feature space is learned in the feature reconstruction process. Thus, the class prototypes for all source and target domains can be constructed in the extracted feature space based on the previous domain prototypes and inherent class prototypes. By forcing the extracted features are close to the corresponding class prototypes for all domains, the feature extraction network is progressively adjusted. In the end, the inherent class prototypes are used as a classifier in the target domain. Our contribution is that through the inherent class prototypes and domain prototypes, the semantic category information from source domains is transformed into the target domain by constructing the corresponding class prototypes. In our method, all source and target domains are aligned twice at the feature level for better domain-invariant features and more closer features to the class prototypes, respectively. Several experiments on public data sets also prove the effectiveness of our method.

Transfer Learning for Remaining Useful Life Prediction Across Operating Conditions Based on Multisource Domain Adaptation

Nowadays, transfer learning and domain adaptation are widely used in prognostics and health management of rotating machinery, greatly broadening its applications in scenarios with multiple operating conditions. However, almost all existing prognostic methods employ a single-source domain adaptation (SDA), ignoring the domain-shift within the source domain, and fail to make full use of multi-source historical data. To overcome these drawbacks, a multisource domain adaptation network (MDAN) for regression tasks is proposed in this article. MDAN learns domain-invariant features and supervision from multiple sources and provides better generalization in the target domain than traditional SDA methods. Besides, a new MDAN-based framework is presented for remaining useful life prediction of bearings under multiple operating conditions. Specifically, we input the time-frequency representation of the vibration signals into MDAN to achieve end-to-end training and inference across operating conditions. Case studies on the run-to-failure datasets validate the effectiveness of proposed method, and the comparison shows that it outperforms the state-of-the-art methods.

A dual alignment-based multi-source domain adaptation framework for motor imagery EEG classification.

Domain adaptation, as an important branch of transfer learning, can be applied to cope with data insufficiency and high subject variabilities in motor imagery electroencephalogram (MI-EEG) based brain-computer interfaces. The existing methods generally focus on aligning data and feature distribution; however, aligning each source domain with the informative samples of the target domain and seeking the most appropriate source domains to enhance the classification effect has not been considered. In this paper, we propose a dual alignment-based multi-source domain adaptation framework, denoted DAMSDAF. Based on continuous wavelet transform, all channels of MI-EEG signals are converted respectively and the generated time-frequency spectrum images are stitched to construct multi-source domains and target domain. Then, the informative samples close to the decision boundary are found in the target domain by using entropy, and they are employed to align and reassign each source domain with normalized mutual information. Furthermore, a multi-branch deep network (MBDN) is designed, and the maximum mean discrepancy is embedded in each branch to realign the specific feature distribution. Each branch is separately trained by an aligned source domain, and all the single branch transfer accuracies are arranged in descending order and utilized for weighted prediction of MBDN. Therefore, the most suitable number of source domains with top weights can be automatically determined. Extensive experiments are conducted based on 3 public MI-EEG datasets. DAMSDAF achieves the classification accuracies of 92.56%, 69.45% and 89.57%, and the statistical analysis is performed by the kappa value and t-test. Experimental results show that DAMSDAF significantly improves the transfer effects compared to the present methods, indicating that dual alignment can sufficiently use the different weighted samples and even source domains at different levels as well as realizing optimal selection of multi-source domains.

Cross-domain knowledge distillation for text classification

Most text classification methods achieve great success based on the large-scale annotated data and the pre-trained language models. However, the labeled data is insufficient in practice, and the pre-trained language models are difficult to deploy due to their high computing resources and slow inference speed. In this paper, we propose cross-domain knowledge distillation, where the teacher and student tasks belong to different domains. It not only acquires knowledge from multiple teachers but also accelerates inference and reduces model size. Specifically, we train the pre-trained language models on factual knowledge obtained by aligned Wikipedia text to Wikidata triplets and fine-tune it as the teacher model. Then we use the heterogeneous multi-teacher knowledge distillation to transfer knowledge from the multiple teacher models to the student model. Multi-teacher knowledge vote can distill knowledge related to the target domain. Moreover, we also introduce the teacher assistant to help distill large pre-trained language models. Finally, we reduce the difference between the source domain and target domain by multi-source domain adaptation to solve the domain shift problem. Experiments on the multiple public datasets demonstrate that our method can achieve competitive performance while having fewer parameters and less inference time.

Unsupervised multi-source domain adaptation with graph convolution network and multi-alignment in mixed latent space

This paper proposes an unsupervised multi-source domain adaptation algorithm with graph convolution network and multi-alignment in mixed latent space, which leverages domain labels, data structure, and category labels in a unified network but improves domain-invariant semantic representation by several innovations. Specifically, a novel data structure alignment is proposed to exploit the inherent properties of different domains while using current domain alignment and classification result alignment. Through this design, category consistency can be considered in both latent space, and domain and structure discrepancy between different source domains and the target domain can be eliminated. Moreover, we also use category alignment based on both CNN and GCN features to optimize category decision boundary. Experiment results show that the proposed method brings sufficient improvement especially for adaptation tasks with large shift in data distribution.