Affective brain-computer interface based on electroencephalography (EEG) is an important branch in the field of affective computing. However, the individual differences in EEG emotional data and the noisy labeling problem in the subjective feedback seriously limit the effectiveness and generalizability of existing models. To tackle these two critical issues, we propose a novel transfer learning framework with Prototypical Representation based Pairwise Learning ( PR-PL ). The discriminative and generalized EEG features are learned for emotion revealing across individuals and the emotion recognition task is formulated as pairwise learning for improving the model tolerance to the noisy labels. More specifically, a prototypical learning is developed to encode the inherent emotion-related semantic structure of EEG data and align the individuals' EEG features to a shared common feature space under consideration of the feature separability of both source and target domains. Based on the aligned feature representations, pairwise learning with an adaptive pseudo labeling method is introduced to encode the proximity relationships among samples and alleviate the label noises effect on modeling. Extensive results on two benchmark databases (SEED and SEED-IV) under four different cross-validation evaluation protocols validate the model reliability and stability across subjects and sessions. Compared to the literature, the average enhancement of emotion recognition across four different evaluation protocols is 2.04% (SEED) and 2.58% (SEED-IV). The source code is available at https://github.com/KAZABANA/PR-PL .

Dynamic-inner canonical correlation analysis (DiCCA) extracts dynamic latent variables from high-dimensional time series data with a descending order of predictability in terms of $R^2$ . The reduced dimensional latent variables with rank-ordered predictability capture the dynamic features in the data, leading to easy interpretation and visualization. In this article, numerically efficient algorithms for DiCCA are developed to extract dynamic latent components from high-dimensional time series data. The numerically improved DiCCA algorithms avoid repeatedly inverting a covariance matrix inside the iteration loop of the numerical DiCCA algorithms. A further improvement using singular value decomposition converts the generalized eigenvector problem into a standard eigenvector problem for the DiCCA solution. Another improvement in model efficiency in this article is the dynamic model compaction of the extracted latent scores using autoregressive integrated moving average (ARIMA) models. Integrating factors, if existed in the latent variable scores, are made explicit in the ARIMA models. Numerical tests on two industrial datasets are provided to illustrate the improvements.

The research of intelligent fault diagnosis method has made great progress. The prerequisite for the effectiveness of most intelligent diagnosis models is to have abundant labeled data, which is difficult to satisfy in practice. Fortunately, we can obtain a large amount of rolling bearing failure data under laboratory conditions. Inspired by the idea of transfer learning, we propose a deep dynamic adaptive transfer network (DDATN) for intelligent fault diagnosis of rolling bearings. In addition to performing transfer diagnosis under different working conditions and failure degrees of the same type of bearing, it is also able to accomplish the task of cross-machine fault diagnosis from bearings under laboratory conditions to the bearings in practical applications. In the DDATN, the marginal probability distribution and conditional probability distribution of the data are aligned by dynamic domain adaptation using weight factor. First, the original vibration signal of the bearing is first processed to establish the source and target domains. Then, pseudo-label learning on target domain unlabeled data is performed and the transferable features between domains are extracted through the deep parameter-shared neural networks. Next, by performing dynamic adaptation on the extracted transferable features and optimizing the intelligent fault diagnosis model through backpropagation, the complete transfer diagnosis task in the target domain is accomplished. The effectiveness of the proposed DDATN method is demonstrated through variable working conditions and cross-machine transfer fault diagnosis tasks. Compared with other intelligent fault diagnosis methods, the proposed method shows clear advantages.

Sufficient data is necessary for intelligent fault diagnostic approaches. However, in practice, it is often the case that only limited fault data is available due to various reasons, making it a challenge to accurately identify the health condition of bearings. To deal with the limited fault data issue, data augmentation strategies, such as generative adversarial network (GAN), are widely utilized. However, GAN has the disadvantages of being difficult to train and restricted ability to generate new data when the fault sample size is limited. Specifically, GAN requires a long training time and abundant training data to make the distribution of generated data closer to the distribution of actual data. This article presents a novel data augmentation approach with compressed sensing for fault diagnosis of bearings to better address the issue of limited fault data. The generated data by compressed sensing is diverse. In addition, the generated data is highly similar to the original data in frequency domain, thus retaining the main feature information of the original data. Furthermore, data augmentation achieved through compressed sensing requires less training data and has lower computational complexity. For bearing fault diagnosis under limited failure data, the limited fault data is first augmented based on compressed sensing, allowing for high fidelity reconstruction and high diversity data generation. Then, the augmented data is utilized to train a deep convolutional neural network to automatically learn and extract features for fault identification. The effectiveness of the presented approach is verified using two bearing datasets.

Intelligent transfer domain adaptation fault diagnosis approaches have received more and more attention. An important prerequisite for the effectiveness of the existing transfer diagnosis methods is to have complete target machine data of all failure types for domain adaptation. However, in practical applications, it is often difficult and time-consuming to obtain all kinds of target fault data. In addition, the diagnosis model cannot be updated to adapt to the actual operating conditions. To overcome these limitations, we propose a time-varying online transfer learning (TVOTL) model for intelligent fault diagnosis of rolling bearings, which can not only construct the transfer diagnosis model with incomplete unlabeled target data offline but also update dynamically during online application. In the offline stage, the initial feature transformation matrix is first obtained using limited inter-domain data and the raw data are mapped to the corresponding feature space to get new representations. Then, in the online stage, data distribution discrepancies between domains are further reduced with online collected target instances and the transfer diagnosis model is updated. In the case study, the proposed TVOTL model is tested on ten transfer diagnosis tasks to illustrate the outstanding performance compared with other online fault diagnosis models.

Electrocardiography (ECG) is a non-invasive tool for predicting cardiovascular diseases (CVDs). Current ECG-based diagnosis systems show promising performance owing to the rapid development of deep learning techniques. However, the label scarcity problem, the co-occurrence of multiple CVDs and the poor performance on unseen datasets greatly hinder the widespread application of deep learning-based models. Addressing them in a unified framework remains a significant challenge. To this end, we propose a multi-label semi-supervised model (ECGMatch) to recognize multiple CVDs simultaneously with limited supervision. In the ECGMatch, an ECGAugment module is developed for weak and strong ECG data augmentation, which generates diverse samples for model training. Subsequently, a hyperparameter-efficient framework with neighbor agreement modeling and knowledge distillation is designed for pseudo-label generation and refinement, which mitigates the label scarcity problem. Finally, a label correlation alignment module is proposed to capture the co-occurrence information of different CVDs within labeled samples and propagate this information to unlabeled samples. Extensive experiments on four datasets and three protocols demonstrate the effectiveness and stability of the proposed model, especially on unseen datasets. As such, this model can pave the way for diagnostic systems that achieve robust performance on multi-label CVDs prediction with limited supervision.