Mean Discrepancy Research Articles

Contrastive learning-enabled digital twin framework for fault diagnosis of rolling bearing

Abstract Rolling bearings are essential components in various industrial machines, and their failures can lead to significant downtime and maintenance costs. Traditional data-driven fault diagnosis methods often require extensive fault datasets for training, which may not always be available in critical industrial scenarios, limiting their practicality. Digital twins, virtual representations of physical entities reflecting their operational conditions, offer a promising solution for the fault diagnosis of rolling bearings with limited fault data. In this paper, we propose a novel digital twin-driven framework to address the challenge of limited training data in rolling bearing fault diagnosis. Firstly, a virtual bearing simulation model is used to generate the simulated data. Subsequently, a transformer-based network is introduced to learn the discrepancy features from the raw data. Then, a Maximum Mean Discrepancy loss and a supervised contrastive learning loss for raw and augmentation data are established to achieve global domain alignment and instance-based domain alignment. Finally, an unsupervised contrastive learning loss for the augmentation data of the target domain is established to further improve the diagnostic performance. In five cross-domain fault diagnosis tasks representing real industrial scenarios set in this study, the proposed method achieved an average diagnostic accuracy of 84.40%, exceeding two existing advanced domain adaptation methods by more than 10%. The experimental results demonstrate that the proposed method achieves high diagnostic performance in real industrial scenarios where labeled data is lacking. This shows its significant benefits for monitoring the condition of critical bearings.

Diversity-Representativeness Replay and Knowledge Alignment for Lifelong Vehicle Re-identification

Lifelong vehicle re-identification (LVReID) aims to match a target vehicle across multiple cameras, considering non-stationary and continuous data streams, which fits the needs of the practical application better than traditional vehicle re-identification. Nonetheless, this area has received relatively little attention. Recently, methods for lifelong person re-identification (LPReID) have been emerging, with replay-based methods achieving the best results by storing a small number of instances from previous tasks for retraining, thus effectively reducing catastrophic forgetting. However, these methods cannot be directly applied to LVReID because they fail to simultaneously consider the diversity and representativeness of replayed data, resulting in biases between the subset stored in the memory buffer and the original data. They randomly sample classes, which may not adequately represent the distribution of the original data. Additionally, these methods fail to consider the rich variation in instances of the same vehicle class due to factors such as vehicle orientation and lighting conditions. Therefore, preserving more informative classes and instances for replay helps maintain information from previous tasks and may mitigate the model’s forgetting of old knowledge. In view of this, we propose a novel Diversity-Representativeness Dual-Stage Sampling Replay (DDSR) strategy for LVReID that constructs an effective memory buffer through two stages, i.e. , Cluster-Centric Class Selection and Diverse Instance Mining. Specifically, we first perform class-level sampling based on density in the clustered class-centered feature space and then further mine the diverse, high-quality instances within the selected classes. In addition, we introduce Maximum Mean Discrepancy loss to align the feature distribution between replay data and the new arrivals and apply L2 regularisation in the parameter space to facilitate knowledge transfer, thus enhancing the model’s generalization ability to new tasks. Extensive experiments demonstrate effective improvements of our method compared to current state-of-the-art lifelong ReID methods on the VeRi-776, VehicleID, and VERI-Wild datasets.

Reliable Augmentation and Precise Identification of EPG Waveforms Based on Multi-Criteria DCGAN

The electrical penetration graph (EPG) technique is of great significance in elucidating the mechanisms of virus transmission by piercing-sucking insects and crop resistance to these insects. The traditional method of manually processing EPG signals encounters the drawbacks of inefficiency and subjectivity. This study investigated the data augmentation and automatic identification of various EPG signals, including A, B, C, PD, E1, E2, and G, which correspond to distinct behaviors exhibited by the Asian citrus psyllid. Specifically, a data augmentation method based on an improved deep convolutional generative adversarial network (DCGAN) was proposed to address the challenge of insufficient E1 waveforms. A multi-criteria evaluation framework was constructed, leveraging maximum mean discrepancy (MMD) to evaluate the similarity between the generated and real data, and singular value decomposition (SVD) was incorporated to optimize the training iterations of DCGAN and ensure data diversity. Four models, convolutional neural network (CNN), K-nearest neighbors (KNN), decision tree (DT), and support vector machine (SVM), were established based on DCGAN to classify the EPG waveforms. The results showed that the parameter-optimized DCGAN strategy significantly improved the model accuracies by 5.8%, 6.9%, 7.1%, and 7.9% for DT, SVM, KNN, and CNN, respectively. Notably, DCGAN-CNN effectively addressed the skewed distribution of EPG waveforms, achieving an optimal classification accuracy of 94.13%. The multi-criteria optimized DCGAN-CNN model proposed in this study enables reliable augmentation and precise automatic identification of EPG waveforms, holding significant practical implications for understanding psyllid behavior and controlling citrus huanglongbing.

Deep joint subdomain alignment for unsupervised domain adaptation

Domain adaptation (DA) aims to transfer knowledge of the labeled source domain to the unlabeled target domain. Current DA methods learn domain-invariant features by aligning source and target feature spaces by distribution discrepancy minimization or adversarial training. However, existing DA methods fail to consider the semantic-level alignment, which can lead to the distortion of semantic feature structures for the target tasks. In this paper, we propose a novel domain adaptation method, called Deep Joint Subdomain Alignment (DJSA), which jointly conducts intra-subdomain and inter-subdomain alignment for fine-grained classification of target tasks. DJSA obtains the pseudo labels of unlabeled target samples by measuring the similarity between the target samples and source category centers, and then DJSA calculates an intra-subdomain loss to reduce the difference within the same class and increase the difference between different classes in both the source domain and the target domain, and DJSA utilizes a Local Maximum Mean Discrepancy (LMMD) metric to measure the inter-subdomain discrepancy between the source and target domain, and minimizes the discrepancy to align different domains. Furthermore, extensive experimental results demonstrate the effectiveness and superiority of our proposed method in unsupervised domain adaptation.

Transfer learning and source domain restructuring-based BiLSTM approach for building energy consumption prediction

ABSTRACT Currently, building energy consumption prediction typically relies on vast amounts of historical data. However, for newly constructed buildings, the scarcity of data leads to reduced prediction accuracy. To address this challenge, this paper proposes a novel approach that integrates transfer learning with a source domain reconstruction-based BiLSTM model for building energy consumption prediction. In the first stage, both source and target domains are clustered into profile types using k-means. For each profile type in the target domain, the most similar profiles in the source domain are identified using Maximum Mean Discrepancy and Dynamic Time Warping. The source domain is then reconstructed by combining these identified profiles based on their proportions in the target domain. Subsequently, a feature extraction method based on EMD-CWT-Conv is introduced. Empirical Mode Decomposition is first applied to decompose and filter the source domain data. Continuous Wavelet Transform is then employed to extract distinctive frequency-domain and time-domain features from both the reconstructed source domain and the target domain. Final predictions are made using transfer learning and fine-tuning. Experiments on a grocery shop and a school show that the proposed approach reduces Mean Absolute Percentage Error by at least 13.19% and 17.67%, respectively.

An Improving Lung Disease Detection by Combining Ensemble Deep Learning and Maximum Mean Discrepancy Transfer Learning

An Improving Lung Disease Detection by Combining Ensemble Deep Learning and Maximum Mean Discrepancy Transfer Learning

A Transferable Meta-Learning Phase Prediction Model for High-Entropy Alloys Based on Adaptive Migration Walrus Optimizer

The phases of high-entropy alloys (HEAs) are crucial to their material properties. Although meta-learning can recommend a desirable algorithm for materials designers, it does not utilize the optimal solution information of similar historical problems in the HEA field. To address this issue, a transferable meta-learning model (MTL-AMWO) based on an adaptive migration walrus optimizer is proposed to predict the phases of HEAs. Firstly, a transferable meta-learning algorithm frame is proposed, which consists of meta-learning based on adaptive migration walrus optimizer, balanced-relative density peaks clustering, and transfer strategy. Secondly, an adaptive migration walrus optimizer model is proposed, which adaptively migrates walruses according to the changes in the average fitness value of the population over multiple iterations. Thirdly, balanced-relative density peaks clustering is proposed to cluster the samples in the source and target domains into several clusters with similar distributions, respectively. Finally, the transfer strategy adopts the maximum mean discrepancy to find the most matching historical problem and transfer its optimal solution information to the target domain. The effectiveness of MTL-AMWO is validated on 986 samples from six datasets, including 323 quinary HEAs, 366 senary HEAs, and 297 septenary HEAs. The experimental results show that the MTL-AMWO achieves better performance than other algorithms.

Dynamic liquid level prediction for multiple oil wells based on transfer learning and multidimensional feature fusion network

Abstract Accurate prediction of the dynamic liquid level (DLL) in oil wells is crucial for the intelligent optimization of pumping systems. It not only provides real-time insights into the operational conditions of the pumping system but also supports the optimization of operational parameters with data. However, due to the long-term operation of oil wells and their complex internal environments, direct measurement of the DLL is challenging, leading to low reliability of the obtained data. Therefore, this paper conducts an in-depth analysis of the parameters involved in the pumping process, identifies the model's input features, and develops a DLL prediction model for multiple wells based on multidimensional feature fusion (MFF). This model captures the characteristics of DLL changes and the diversity of input features. To address the issues of slow model training and low prediction accuracy caused by insufficient datasets in practical applications, this paper integrates transfer learning techniques and proposes a new model, the DLL model for multiple wells based on Transfer Learning and Multidimensional Feature Fusion (TMFF). Initially, the Euclidean distance and Maximum mean discrepancy methods are employed to verify the feature similarity between the source and target domains, using highly similar DLL data as experimental data. By combining transfer learning techniques with the MFF model, the TMFF model is established. The model's capabilities are validated using field-collected data with broad representativeness. Experimental results demonstrate that the proposed MFF model possesses high accuracy and generalization capability. Additionally, the TMFF model effectively resolves the issue of insufficient data during model training. In summary, the methods proposed in this paper can provide accurate DLL data for practical applications in intelligent oilfields.

A Novel Fault Diagnosis Method for Rolling Bearings Based on Multi-Domain Adaptation Neural Network

In recent years, data-driven approaches have shown promise in fault diagnosis. However, in practical industrial applications, challenges such as discrepancies between source domain and target domain data distributions, coupled with limited labeled fault data, have hindered the performance of traditional domain adaptation algorithms in bearing fault diagnosis. To address this issue, this study introduces a novel method based on MDANN (Multi Domain Adaptation Neural Network) for diagnosing rolling bearing faults using unlabeled data. Initially, the raw vibration signals are preprocessed using wavelet packet decomposition and reconstruction (WPT), which helps minimize signal redundancy while preserving critical features. Following this, the multi-kernel maximum mean discrepancy (MK-MMD) algorithm is employed to compute the differences in input feature values, and the MDANN network parameters are adjusted via backpropagation, allowing the network to extract domain-invariant features. To ensure the unlabeled target domain data can be effectively used in training, a pseudo-labeling strategy is applied, where the label with the highest probability is treated as the true label, thereby enabling the model to learn from the target domain data and improve the acquisition of reliable diagnostic knowledge. The method is validated using two publicly available datasets, CWRU and PU, with experimental results demonstrating that it outperforms conventional domain adaptation techniques in terms of diagnostic accuracy. This indicates the proposed approach is highly effective in capturing transferable features and addressing the distributional differences between datasets.

Realtime and noninvasive assessment of endotracheal tube displacement using near-infrared and visible cameras

Endotracheal tube (ETT) intubation is a medical procedure routinely used for achieving mechanical ventilation in critically ill patients. Appropriate ETT placement is crucial as undetected tube migration may cause multiple complications or even fatalities. Therefore, prompt detection of unplanned movement of the ETT and immediate action to restore proper placement are essential to ensure patient safety. Despite this necessity, there is not a widely adopted tool for real-time assessment of ETT displacement. We have developed a device, a dual-camera endotracheal tube or DC-ETT, to address this unmet clinical need. This device uses a near-infrared (NIR) LED and a side-firing optical fiber embedded in the side of an ETT to light up the tracheal tissue and a visible and NIR camera module for the displacement detection. The NIR camera tracks the movement of the NIR pattern on the skin, while the visible camera is used to correct the body movements. The efficacy of the DC-ETT was assessed in two piglets with a linear displacement sensor as reference. A mean discrepancy of less than 0.5 mm between the DC-ETT and reference sensor was observed within a displacement range of ±15 mm. The results suggest that the DC-ETT can potentially provide a simple and cost-effective solution for real-time monitoring of ETT displacements in operating rooms, intensive care units, and emergency departments.

Multi-UAV Cooperative Pursuit of a Fast-Moving Target UAV Based on the GM-TD3 Algorithm

Recently, developing multi-UAVs to cooperatively pursue a fast-moving target has become a research hotspot in the current world. Although deep reinforcement learning (DRL) has made a lot of achievements in the UAV pursuit game, there are still some problems such as high-dimensional parameter space, the ease of falling into local optimization, the long training time, and the low task success rate. To solve the above-mentioned issues, we propose an improved twin delayed deep deterministic policy gradient algorithm combining the genetic algorithm and maximum mean discrepancy method (GM-TD3) for multi-UAV cooperative pursuit of high-speed targets. Firstly, this paper combines GA-based evolutionary strategies with TD3 to generate action networks. Then, in order to avoid local optimization in the algorithm training process, the maximum mean difference (MMD) method is used to increase the diversity of the policy population in the updating process of the population parameters. Finally, by setting the sensitivity weights of the genetic memory buffer of UAV individuals, the mutation operator is improved to enhance the stability of the algorithm. In addition, this paper designs a hybrid reward function to accelerate the convergence speed of training. Through simulation experiments, we have verified that the training efficiency of the improved algorithm has been greatly improved, which can achieve faster convergence; the successful rate of the task has reached 95%, and further validated UAVs can better cooperate to complete the pursuit game task.

An Evidential Multi-Target Domain Adaptation Method Based on Weighted Fusion for Cross-Domain Pattern Classification.

For cross-domain pattern classification, the supervised information (i.e., labeled patterns) in the source domain is often employed to help classify the unlabeled target domain patterns. In practice, multiple target domains are usually available. The unlabeled patterns (in different target domains) which have high-confidence predictions, can also provide some pseudo-supervised information for the downstream classification task. The performance in each target domain would be further improved if the pseudo-supervised information in different target domains can be effectively used. To this end, we propose an evidential multi-target domain adaptation (EMDA) method to take full advantage of the useful information in the single-source and multiple target domains. In EMDA, we first align distributions of the source and target domains by reducing maximum mean discrepancy (MMD) and covariance difference across domains. After that, we use the classifier learned by the labeled source domain data to classify query patterns in the target domains. The query patterns with high-confidence predictions are then selected to train a new classifier for yielding an extra piece of soft classification results of query patterns. The two pieces of soft classification results are then combined by evidence theory. In practice, their reliabilities/weights are usually diverse, and an equal treatment of them often yields the unreliable combination result. Thus, we propose to use the distribution discrepancy across domains to estimate their weighting factors, and discount them before fusing. The evidential combination of the two pieces of discounted soft classification results is employed to make the final class decision. The effectiveness of EMDA was verified by comparing with many advanced domain adaptation methods on several cross-domain pattern classification benchmark datasets.

GAN-based statistical process control for the time series data

The cumulative sum (CUSUM) control chart and the multivariate anomaly detection with the generative adversarial network (MAD-GAN) were compared for monitoring the time series data. However, the control boundaries constructed in terms of the one-class classification with only the normal data for the training phase are inappropriate for the test phase because the normal data and the abnormal data should be classified for the test phase. In this regard, we first propose this GAN-based statistical process control (SPC) framework to compare them in terms of detecting the process mean shift based on the perspective of SPC. Second, we propose the residual MAD-GAN in order to improve the detection performance. Third, we develop the loss function of the MAD-GAN. Finally, we find that the maximum mean discrepancy (MMD) as well as the nash equilibrium is useful for the MAD-GAN. Our experiments demonstrate that the residual MAD-GAN is more effective than the residual CUSUM control chart in terms of the run lengths for the time series data. Therefore, we propose SPC practitioners to consider the residual MAD-GAN for detecting the process mean shift in time series data.

Testing the equality of distributions using integrated maximum mean discrepancy

Comparing and testing for the homogeneity of two independent random samples is a fundamental statistical problem with many applications across various fields. However, existing methods may not be effective when the data is complex or high-dimensional. We propose a new method that integrates the maximum mean discrepancy (MMD) with a Gaussian kernel over all one-dimensional projections of the data. We derive the closed-form expression of the integrated MMD and prove its validity as a distributional similarity metric. We estimate the integrated MMD with the U-statistic theory and study its asymptotic behaviors under the null and two kinds of alternative hypotheses. We demonstrate that our method has the benefits of the MMD, and outperforms existing methods on both synthetic and real datasets, especially when the data is complex and high-dimensional.

Rigid-flexible coupling dynamic-assisted imbalanced fault diagnosis for helicopter tail transmission system

Helicopters operate in extreme environments, complicating the balanced and accurate acquisition of operational data. Predicting high-precision failures in helicopter tail-drive systems is particularly challenging due to severe data imbalances. To address this, a fault diagnosis method based on rigid-flexible coupling dynamics is proposed. A high-precision simulation model generates source domain data samples, validated by comparing the frequency domain components and RMS values of simulated and tested vibration responses under healthy conditions. An adaptive Local Maximum Mean Discrepancy (LMMD) mechanism aligns features between simulated and tested data. Additionally, the Coordinate-Separable Convolutional ResNet (CSC-ResNet) network is proposed to efficiently learn high-dimensional feature knowledge. Results show the proposed method outperforms others in fault diagnosis accuracy and smoothness, offering a new perspective for fault diagnosis of helicopter tail transmission systems.

Cross-Subject Seizure Detection via Unsupervised Domain-Adaptation.

Automatic seizure detection from Electroencephalography (EEG) is of great importance in aiding the diagnosis and treatment of epilepsy due to the advantages of convenience and economy. Existing seizure detection methods are usually patient-specific, the training and testing are carried out on the same patient, limiting their scalability to other patients. To address this issue, we propose a cross-subject seizure detection method via unsupervised domain adaptation. The proposed method aims to obtain seizure specific information through shallow and deep feature alignments. For shallow feature alignment, we use convolutional neural network (CNN) to extract seizure-related features. The distribution gap of the shallow features between different patients is minimized by multi-kernel maximum mean discrepancies (MK-MMD). For deep feature alignment, adversarial learning is utilized. The feature extractor tries to learn feature representations that try to confuse the domain classifier, making the extracted deep features more generalizable to new patients. The performance of our method is evaluated on the CHB-MIT and Siena databases in epoch-based experiments. Additionally, event-based experiments are also conducted on the CHB-MIT dataset. The results validate the feasibility of our method in diminishing the domain disparities among different patients.

Dynamic robustness evaluation for automated model selection in operation

Context:The increasing use of artificial neural network (ANN) classifiers in systems, especially safety-critical systems (SCSs), requires ensuring their robustness against out-of-distribution (OOD) shifts in operation, which are changes in the underlying data distribution from the data training the classifier. However, measuring the robustness of classifiers in operation with only unlabeled data is challenging. Additionally, machine learning engineers may need to compare different models or versions of the same model and switch to an optimal version based on their robustness. Objective:This paper explores the problem of dynamic robustness evaluation for automated model selection. We aim to find efficient and effective metrics for evaluating and comparing the robustness of multiple ANN classifiers using unlabeled operational data. Methods:To quantitatively measure the differences between the model outputs and assess robustness under OOD shifts using unlabeled data, we choose distance-based metrics. An empirical comparison of five such metrics, suitable for higher-dimensional data like images, is performed. The selected metrics include Wasserstein distance (WD), maximum mean discrepancy (MMD), Hellinger distance (HL), Kolmogorov–Smirnov statistic (KS), and Kullback–Leibler divergence (KL), known for their efficacy in quantifying distribution differences. We evaluate these metrics on 20 state-of-the-art models (ten CIFAR10-based models, five CIFAR100-based models, and five ImageNet-based models) from a widely used robustness benchmark (RobustBench) using data perturbed with various types and magnitudes of corruptions to mimic real-world OOD shifts. Results:Our findings reveal that the WD metric outperforms others when ranking multiple ANN models for CIFAR10- and CIFAR100-based models, while the KS metric demonstrates superior performance for ImageNet-based models. MMD can be used as a reliable second option for both datasets. Conclusion:This study highlights the effectiveness of distance-based metrics in ranking models’ robustness for automated model selection. It also emphasizes the significance of advancing research in dynamic robustness evaluation.

SDYN-GANs: Adversarial learning methods for multistep generative models for general order stochastic dynamics

We introduce adversarial learning methods for data-driven generative modeling of dynamics of nth-order stochastic systems. Our approach builds on Generative Adversarial Networks (GANs) with generative model classes based on stable m-step stochastic numerical integrators. From observations of trajectory samples, we introduce methods for learning long-time predictors and stable representations of the dynamics. Our approaches use discriminators based on Maximum Mean Discrepancy (MMD), training protocols using both conditional and marginal distributions, and methods for learning dynamic responses over different time-scales. We show how our approaches can be used for modeling physical systems to learn force-laws, damping coefficients, and noise-related parameters. Our adversarial learning approaches provide methods for obtaining stable generative models for dynamic tasks including long-time prediction and developing simulations for stochastic systems.

FireDA: A Domain Adaptation-Based Method for Forest Fire Recognition with Limited Labeled Scenarios

Vision-based forest fire detection systems have significantly advanced through Deep Learning (DL) applications. However, DL-based models typically require large-scale labeled datasets for effective training, where the quality of data annotation is crucial to their performance. To address challenges related to the quality and quantity of labeling, a domain adaptation-based approach called FireDA is proposed for forest fire recognition in scenarios with limited labels. Domain adaptation, a subfield of transfer learning, facilitates the transfer of knowledge from a labeled source domain to an unlabeled target domain. The construction of the source domain FBD is initiated, which includes three common fire scenarios: forest (F), brightness (B), and darkness (D), utilizing publicly available labeled data. Subsequently, a novel algorithm called Neighborhood Aggregation-based 2-Stage Domain Adaptation (NA2SDA) is proposed. This method integrates feature distribution alignment with target domain Proxy Classification Loss (PCL), leveraging a neighborhood aggregation mechanism and a memory bank designed for the unlabeled samples in the target domain. This mechanism calibrates the source classifier and generates more accurate pseudo-labels for the unlabeled sample. Consequently, based on these pseudo-labels, the Local Maximum Mean Discrepancy (LMMD) and the Proxy Classification Loss (PCL) are computed. To validate the efficacy of the proposed method, the publicly available forest fire dataset, FLAME, is employed as the target domain for constructing a transfer learning task. The results demonstrate that our method achieves performance comparable to the supervised Convolutional Neural Network (CNN)-based state-of-the-art (SOTA) method, without requiring access to labels from the FLAME training set. Therefore, our study presents a viable solution for forest fire recognition in scenarios with limited labeling and establishes a high-accuracy benchmark for future research.

Robust knowledge transfer for bearing diagnosis in neural network models using multilayer maximum mean discrepancy loss function

Abstract This paper introduces a robust transfer learning method to enhance bearing diagnosis, particularly in cross-machine scenarios. The method trains a shallow neural network using labeled data from a different machine and unlabeled data from the target monitoring machine. To facilitate effective knowledge transfer, a multilayer maximum mean discrepancy loss function is employed, enabling the model to adapt learned features from the source machine to the target machine’s unlabeled data. This approach addresses the challenges of low accuracy and robustness often seen in transfer learning, especially when dealing with different machines. Experiments conducted on the Hanoi University of Science and Technology bearing dataset validate the effectiveness of the proposed method. The results show significant improvements in prediction accuracy and robustness, making this method superior to existing transfer learning models for cross-machine bearing diagnosis.