Class Boundaries Research Articles

Digital-analog quantum learning on Rydberg atom arrays

Abstract We propose hybrid digital-analog learning algorithms on Rydberg atom arrays, combining the potentially practical utility and near-term realizability of quantum learning with the rapidly scaling architectures of neutral atoms. Our construction requires only single-qubit operations in the digital setting and global driving according to the Rydberg Hamiltonian in the analog setting. We perform a comprehensive numerical study of our algorithm on both classical and quantum data, given respectively by handwritten digit classification and unsupervised quantum phase boundary learning. We show in the two representative problems that digital-analog learning is not only feasible in the near term, but also requires shorter circuit depths and is more robust to realistic error models as compared to digital learning schemes. Our results suggest that digital-analog learning opens a promising path towards improved variational quantum learning experiments in the near term.

A class alignment network based on self-attention for cross-subject EEG classification

Abstract Due to the inherent variability in EEG signals across different individuals, domain adaptation and adversarial learning strategies are being progressively utilized to develop subject-specific classification models by leveraging data from other subjects. These approaches primarily focus on domain alignment and tend to overlook the critical task-specific class boundaries. This oversight can result in weak correlation between the extracted features and categories. To address these challenges, we propose a novel model that uses the known information from multiple subjects to bolster EEG classification for an individual subject through adversarial learning strategies. Our method begins by extracting both shallow and attention-driven deep features from EEG signals. Subsequently, we employ a class discriminator to encourage the same-class features from different domains to converge while ensuring that the different-class features diverge. This is achieved using our proposed discrimination loss function, which is designed to minimize the feature distance for samples of the same class across different domains while maximizing it for those from different classes. Additionally, our model incorporates two parallel classifiers that are harmonious yet distinct and jointly contribute to decision-making. Extensive testing on two publicly available EEG datasets has validated our model’s efficacy and superiority.

Attention-disentangled re-ID network for unsupervised domain adaptive person re-identification

Unsupervised domain adaptation (UDA) for person re-identification (re-ID) aims to bridge the domain gap by transferring knowledge from the labeled source domain to the unlabeled target domain. Recently, pseudo-label-based approaches have become the dominant solution for addressing this issue. However,most pseudo-label-based methods neglect class boundary samples to reduce the influence of false pseudo labels, sacrificing intra-class semantic diversity. Although hard sample memory bank-based approaches can discover relationships between samples and describe intra-class diversity, they are prone to increasing the risks of generating incorrect pseudo labels. Balancing intra-class variety and pseudo-label accuracy in cross-domain person re-ID poses a challenge. In this study, we propose an attention-disentangled re-ID network (ADDNet) to enhance the discriminative ability of re-ID related feature representations, addressing the contradiction between intra-class diversity and pseudo-label accuracy. Unlike most disentanglement learning-based re-ID approaches that focus on separating explanatory factors, we design a spatial attention-disentangled mechanism to separate re-ID related and unrelated weights, enhancing the discriminative ability of cross-domain feature representation. Additionally, a multiple hard sample memory learning strategy is designed to express intra-class diversity of target samples using re-ID related features extracted from ADDNet. Extensive experiments show that ADDNet achieves 46.7% mAP on the Market-to-MSMT cross-domain re-ID task, outperforming state-of-the-art methods by 6.5 points.

Parkinson’s Disease Detection by Using Machine Learning Method based on Local Classification on Class Boundary

Parkinson’s disease (PD) detection has long been an important task in medical intelligence. Recognition methods based on speech signals show great potential in Parkinson’s disease diagnosis. In this paper, based on an efficient machine learning method for Parkinson’s disease detection, we take the use of test data incorporates an efficient Secure Two-Party Computing (S2PC) protocol to protect the privacy of patients. We present two key components, the secure use of data and a local classification methodology, including the description of class boundaries. We conducted experiments on two datasets to validate our proposed method, and the results show well data security protection ability compared to some more sophisticated methods. And the performance of Local Classification on Class Boundary(LCCB) and Hyperplane K-Nearest Neighbor(HKNN) is significantly better than that of both Support Vector Machines(SVM) and Random Forest(RF). When the number of selected features is from 400 to 500, HKNN and LCCB are roughly equal where the accuracy of HKNN is 95.2%, and LCCB has the rate of 94.7%. Then we use Multi-Cluster Feature Selection(MCFS) to analyze and select the important features from D2 dataset. It shows that even if only two features are selected, the boundaries of the two categories are also clear and easy to distinguish.

A semiparametric accelerated failure time-based mixture cure tree

The mixture cure rate model (MCM) is the most widely used model for the analysis of survival data with a cured subgroup. In this context, the most common strategy to model the cure probability is to assume a generalized linear model with a known link function, such as the logit link function. However, the logit model can only capture simple effects of covariates on the cure probability. In this article, we propose a new MCM where the cure probability is modeled using a decision tree-based classifier and the survival distribution of the uncured is modeled using an accelerated failure time structure. To estimate the model parameters, we develop an expectation maximization algorithm. Our simulation study shows that the proposed model performs better in capturing nonlinear classification boundaries when compared to the logit-based MCM and the spline-based MCM. This results in more accurate and precise estimates of the cured probabilities, which in-turn results in improved predictive accuracy of cure. We further show that capturing nonlinear classification boundary also improves the estimation results corresponding to the survival distribution of the uncured subjects. Finally, we apply our proposed model and the EM algorithm to analyze an existing bone marrow transplant data.

An Integrated Spatial Fuzzy‐Based Site Suitability Assessment Framework for Agricultural BMP Placement

ABSTRACTAssigning crisp class boundaries to landscape features can result in the loss of vital information for land evaluation objectives, especially when these boundaries lack clear definitions. This challenge becomes particularly pronounced when land suitability is assessed for implementing agricultural best management practices (BMPs)—conservation measures aimed at reducing the environmental risks of farming activities to aquatic ecosystems while simultaneously achieving water quality and economic objectives. To address the limitations associated with Boolean suitability assessment frameworks, we have introduced an integrated spatial, fuzzy‐based land evaluation framework that considers a range of hydrological and economic determinants for BMP placement. By employing data‐driven fuzzy membership functions and overlay operators, this framework generates a joint suitability index for BMP placement across agricultural watersheds. The application of the proposed framework to the Thames River Watershed in southwestern Ontario, Canada, produced the first joint suitability index of the watershed. Further analysis of the average farm‐level joint suitability scores identified statistically significant clusters of highly suitable and unsuitable lands for BMP placement, with 85% of highly suitable lands being situated in the upper basin areas. The proposed framework is adaptable to various agricultural production geographies, especially in data‐limited environments, allowing for strategic BMP placement to mitigate the global impacts of anthropogenic nutrient loadings on aquatic ecosystems. For optimal results, context‐specific applications should prioritize research on locally relevant fuzzy membership functions and BMP implementation drivers.

Cross-domain transfer fault diagnosis by class-imbalanced deep subdomain adaptive network

In variable working conditions and class imbalances, fault diagnosis in gearbox systems becomes difficult due to the difficulty in learning the classification boundaries of the minority class. Moreover, the diversity in fault mode distributions under various working conditions poses significant challenges for domain adaptation. To this end, this work proposed a Class-imbalanced Deep Subdomain Adaptive Network (CIDSAN) for complex cross-domain fault diagnosis scenarios involving feature and label shifts. We design a convolutional neural network embedded with SimAM (SimAM-CNN) to extract global feature information from time-domain vibration signals. Subsequently, a novel cost-sensitive classifier is presented for class re-weighting, along with a Joint Supervised Classifier that refines intra-class and inter-class distances by re-weighting class weights in the latent space using cost-sensitive classifiers while leveraging LDAM regularization to adjust classification boundaries. Diagnostic case studies on two datasets under various working conditions validate the effectiveness and superiority of the proposed approach. The code https://github.com/Zjy0121/CIDSAN.

LightGBM integration with modified data balancing and whale optimization algorithm for rock mass classification

The accurate prediction of uneven rock mass classes is crucial for intelligent operation in tunnel-boring machine (TBM) tunneling. However, the classification of rock masses presents significant challenges due to the variability and complexity of geological conditions. To address these challenges, this study introduces an innovative predictive model combining the improved EWOA (IEWOA) and the light gradient boosting machine (LightGBM). The proposed IEWOA algorithm incorporates a novel parameter l for more effective position updates during the exploration stage and utilizes sine functions during the exploitation stage to optimize the search process. Additionally, the model integrates a minority class technique enhanced with a random walk strategy (MCT-RW) to extend the boundaries of minority classes, such as Classes II, IV, and V. This approach significantly improves the recall and F1-score for these rock mass classes. The proposed methodology was rigorously evaluated against other predictive algorithms, demonstrating superior performance with an accuracy of 94.74%. This innovative model not only enhances the accuracy of rock mass classification but also contributes significantly to the intelligent and efficient construction of TBM tunnels, providing a robust solution to one of the key challenges in underground engineering.

Multi-reference source aircraft classification network based on unknown class prototype estimation

Abstract Gathering complete aircraft types for close-set tasks is challenging and costly in fine-grained aircraft classification, resulting in encountering unknown class aircraft images in real-world models. To address this problem, we propose a network called a multi-reference source fine-grained aircraft classification network (MRSN) to explicitly and implicitly distinguish known class boundaries and embed unknown classes into the sample space. Specifically, we propose an embedding module (UCPE) to synthesize an unknown class prototype to facilitate the modeling of unknown class data distribution. Besides, we introduce a concatenated multi-reference source module (CMS) that combines distance measurement and probability estimation to classify whether testing images are of unknown class or not. Our proposed MRSN innovatively extends the open-set learning paradigm to the fine-grained classification of geospatial objects rather than simple scene classification. Besides, it shows excellent performance in our constructed dataset, which demonstrates the effectiveness and progressiveness of our method.

Resampling approach for imbalanced data classification based on class instance density per feature value intervals

In practical applications, imbalanced datasets significantly degrade the classification performance of machine learning models. However, most conventional resampling approaches fall short in adequately addressing the varying contributions of individual features to the classification model. In response to this defect, this study introduces three novel resampling approaches. The first approach, Oversampling based on class instance density per feature value intervals (OCF), focuses on augmenting the dataset. The second approach, Undersampling based on class instance density per feature value intervals (UCF), seeks to reduce dataset size. The third approach, Hybrid sampling based on class instance density per feature value intervals (HSCF), which can perform oversampling and undersampling simultaneously. These approaches categorize feature value into different intervals based on their varying information content, calculate class instance densities within these intervals, and generate feature values in intervals with high discriminative information. Subsequently, these generated features are combined to synthesize minority class data, effectively achieving oversampling. Additionally, the study combines class instance density and feature importance to identify majority class data at the classification boundary with minimal contribution and subsequently executes undersampling. The flexibility to adjust sampling ratios and the integration of OCF and UCF enable the implementation of hybrid sampling. Finally, experiments on the benchmark dataset demonstrate the superiority and effectiveness of the proposed method. Furthermore, it is observed that the method proposed in this study enhances the feature dividing capability of decision tree classifiers. Hence, the best results are achieved when working in synergy with decision tree classifiers, leading to the most significant improvements in classification performance. All codes have been published at https://github.com/Wangfeiopen/HSCF.

Anti-forgetting source-free domain adaptation method for machine fault diagnosis

Unsupervised domain adaptation methods have made significant progress in the field of machine fault diagnosis. However, these methods generally assume the accessibility of source domain data during the cross-domain transfer phase. In practice, this assumption is often impractical due to data privacy requirements and the burden of data transmission and storage. Additionally, a key challenge lies in preventing the forgetting of source domain knowledge while performing cross-domain diagnosis in a source-free scenario. To address these issues, this paper proposes an anti-forgetting source-free domain adaptation method for machine fault diagnosis. Specifically, a prototype-based pseudo-label generation strategy, combined with a nearest-neighbor constraint, is employed to deeply explore diagnosis information in the unlabeled target domain data. Then, a fast nuclear-norm maximization constraint is introduced to reduce the density of target domain samples near the classification boundary, thereby decreasing the probability of misdiagnosis. Furthermore, a weighted Fisher regularization is designed to retain the diagnosis information of healthy states inherent in the source domain, thus mitigating information forgetting. Finally, comprehensive experiments are conducted to validate the superiority of our method in source-free cross-domain diagnosis from multiple perspectives, while also demonstrating its anti-forgetting properties.

Multivariate prototype representation for domain-generalized incremental learning

Deep learning models often suffer from catastrophic forgetting when fine-tuned with samples of new classes. This issue becomes even more challenging when there is a domain shift between training and testing data. In this paper, we address the critical yet less explored Domain-Generalized Class-Incremental Learning (DGCIL) task. We propose a DGCIL approach designed to memorize old classes, adapt to new classes, and reliably classify objects from unseen domains. Specifically, our loss formulation maintains classification boundaries while suppressing domain-specific information for each class. Without storing old exemplars, we employ knowledge distillation and estimate the drift of old class prototypes as incremental training progresses. Our prototype representations are based on multivariate Normal distributions, with means and covariances continually adapted to reflect evolving model features, providing effective representations for old classes. We then sample pseudo-features for these old classes from the adapted Normal distributions using Cholesky decomposition. Unlike previous pseudo-feature sampling strategies that rely solely on average mean prototypes, our method captures richer semantic variations. Experiments on several benchmarks demonstrate the superior performance of our method compared to the state of the art.

Singlish in focus: a personal view from Downstairs

ABSTRACT The status of Singlish in Singapore media is a site of conflict between the government and the populace. Linguistic control in Singapore, as a strategy for managing national identity and cohesion, is not new. However, understanding the regulation of Singlish in television and cinema elucidates the media’s role in Singapore’s language policy. Singlish validates the lived experience of Singaporeans in media representations, yet the government discourages its use. This paper examines the language use in the first two seasons of the adult animated series Downstairs (Han) to explore Singlish’s role in local media as a conduit of cohesion, division and transgression. Brief comparisons are made to three prominent movies released in the same period: A Land Imagined (Yeo), A Long Long Time Ago 3: Diam Diam Era and A Long Long Time Ago 4: Diam Diam Era 2 (Neo). Singlish neutralizes identity markers like ethnicity, culture and nationality but emphasizes boundaries of class, education and perceived success. Singlish, as a language of disempowerment and marginalization, symbolizes anti-establishment attitudes and solidarity against English-speaking status. Downstairs demonstrates how language differentiates class, expresses cultural values and addresses social issues, empowering its satire in interrogating the typical Singapore experience.

DNA-based marine benthic assessment methods can perform as morphological ones, but an intercalibration is needed

There is an increasing need for legislation worldwide to monitor and assess the ecological status of marine ecosystems, due to increasing pressures from human activities. The costs and time of traditional analyses are high, while methods based on molecular analysis could reduce these costs and shorten evaluation times significantly. Some biotic indices used to assess the status need reference conditions to be applied. Hence, our objective here is to develop reference conditions for a molecular-based benthic index (M−gAMBI), which can be compared with a morphological one (M−AMBI). Using 6 years of data from estuaries and coasts, we have been able to set reference conditions for five water types, including richness, diversity and AMBI. However, one problem is the absence of the whole human pressure gradient in all water types, making difficult to validate the reference conditions. Based on the results of this research, the M−gAMBI index could be considered suitable for ecological status assessment since it meets most of the criteria for considering a genomics-based index suitable. However, it is considered necessary to advance in (i) improving the detection of errors in genomic methods and similarity with morphological methods, and (ii) an intercalibration exercise, allowing adjusting quality class boundaries and determining the ecological status in an equivalent manner using both methods.

A Novel Active Learning Framework for Cross-Subject Human Activity Recognition from Surface Electromyography.

Wearable sensor-based human activity recognition (HAR) methods hold considerable promise for upper-level control in exoskeleton systems. However, such methods tend to overlook the critical role of data quality and still encounter challenges in cross-subject adaptation. To address this, we propose an active learning framework that integrates the relation network architecture with data sampling techniques. Initially, target data are used to fine tune two auxiliary classifiers of the pre-trained model, thereby establishing subject-specific classification boundaries. Subsequently, we assess the significance of the target data based on classifier discrepancy and partition the data into sample and template sets. Finally, the sampled data and a category clustering algorithm are employed to tune model parameters and optimize template data distribution, respectively. This approach facilitates the adaptation of the model to the target subject, enhancing both accuracy and generalizability. To evaluate the effectiveness of the proposed adaptation framework, we conducted evaluation experiments on a public dataset and a self-constructed electromyography (EMG) dataset. Experimental results demonstrate that our method outperforms the compared methods across all three statistical metrics. Furthermore, ablation experiments highlight the necessity of data screening. Our work underscores the practical feasibility of implementing user-independent HAR methods in exoskeleton control systems.

Self-contrastive Learning-optimized General Agent for long-tailed fault diagnosis of shipboard antennas leveraging adaptive data distribution

To address the challenges of low accuracy and limited generalization in long-tailed fault diagnosis, an adaptive data distribution-based reinforcement learning General Agent is proposed. The method primarily targets more discriminative, domain-invariant feature learning by pre-training the deep Q-network with unlabeled positive samples. Supervisory signals derived from the data’s intrinsic structure enhance class boundary detection while maximizing intra-class feature similarity. Next, empirical data prioritization based on state-action values and TD-error enables efficient utilization of rare but critical experiences, significantly improving sampling efficiency. Concurrently, an adaptive distribution strategy refines a hierarchical reward system by dynamically calibrating the reward function according to real-time accuracy feedback. The deep Q-network, structured with ResNet as the backbone, integrates Efficient Channel Attention (ECA) and Global Attention Mechanism (GAM) to enhance decision-making robustness. Tested on a long-tailed shipboard antenna dataset, the proposed method autonomously identifies fault patterns, demonstrating clear advantages in efficiency, robustness, generalization, and interpretability.

Bidirectional consistency with temporal-aware for semi-supervised time series classification

Semi-supervised learning (SSL) has achieved significant success due to its capacity to alleviate annotation dependencies. Most existing SSL methods utilize pseudo-labeling to propagate useful supervised information for training unlabeled data. However, these methods ignore learning temporal representations, making it challenging to obtain a well-separable feature space for modeling explicit class boundaries. In this work, we propose a semi-supervised Time Series classification framework via Bidirectional Consistency with Temporal-aware (TS-BCT), which regularizes the feature space distribution by learning temporal representations through pseudo-label-guided contrastive learning. Specifically, TS-BCT utilizes time-specific augmentation to transform the entire raw time series into two distinct views, avoiding sampling bias. The pseudo-labels for each view, generated through confidence estimation in the feature space, are then employed to propagate class-related information into unlabeled samples. Subsequently, we introduce a temporal-aware contrastive learning module that learns discriminative temporal-invariant representations. Finally, we design a bidirectional consistency strategy by incorporating pseudo-labels from two distinct views into temporal-aware contrastive learning to construct a class-related contrastive pattern. This strategy enables the model to learn well-separated feature spaces, making class boundaries more discriminative. Extensive experimental results on real-world datasets demonstrate the effectiveness of TS-BCT compared to baselines.

EEG classification with limited data: A deep clustering approach

In this paper, we propose a novel training strategy designed to prevent deep neural networks from overfitting when trained on a very limited number of samples. This approach enables the use of very deep networks in applications like EEG classification, where data collection is challenging, leading to significant overfitting issues. To address this issue, we introduce a classification network that leverages a deep clustering operation in its latent space. The network utilizes a set of dictionary elements as auxiliary inputs and learns to guide each input sample toward one of the clusters, each derived from a dictionary element. This process facilitates the grouping of samples from the same class within the same region of the latent space, thereby reducing the risk of complex class boundaries and overfitting. Additionally, we introduce two probabilistic models within our clustering approach to mitigate outlier issues and address augmentation challenges commonly encountered with EEG signals. Experimental evaluations on two widely recognized databases, EEG-ImageNet and BCIIV2a, demonstrate the effectiveness and superiority of our method compared to state-of-the-art approaches, achieving 97.9% accuracy on EEG-ImageNet and 84.1% accuracy on BCIIV2a.

Analysis and extreme event prediction of waves in the eastern shelf seas of China based on a 44-year hindcast

Analysis and extreme event prediction of regional waves are crucial for the design, spatial layout and protection of ocean engineering structures. In this study, the regional wave characteristic is analysed in the eastern shelf seas of China based on a 44-year wave hindcast. The spatial variation of peak period is smaller than that of significant wave height, which may be affected by swell propagation, and the seasonal distribution differences of these parameters can generally be explained by changes in the East Asian monsoon and are significantly affected by tropical cyclones. Both local wind wave and remotely generated swell should be considered when designing ocean structures. In addition, the joint distribution characteristics of wave period and height reveal a natural regional classification of representative wave states, which can be used to avoid deploying structures near classification boundaries. The extreme wave events are characterized by their number, duration and intensity, highlighting areas susceptible to extreme waves. Meanwhile, the extreme event is predicted, and its return patterns are highly consistent with weather system characteristics such as tropical cyclone tracks. The results can provide a significant basis for the design and protection of ocean engineering structures.

A neural network for recognizing and classifying images at the boundaries of several classes

The neural network belongs to the computer systems of computing technology and artificial intelligence in the field of building automated systems for recognizing and classifying images at the border of several classes, which is made with the ability to perform diagnostics and image recognition at the borders of two, three or more classes due to the fact that the layer schemes for selection of one, two or more single signals from the outputs of the layer of output neurons , , ¼, , ¼, . at each moment of time remembers the number of non-zero signals at the output of neurons of the output layer, and the second layer of output neurons , , ¼, , ¼, remembers neurons of the output layer that have non-zero signals at their outputs. The technical result achieved by this neural network is an increase in the number of classes that can be recognized and the ability to recognize images that are at the same Hamming distance from two, three or more reference images stored in the connection weights of the neurons of the memory layer. haunting An experimental evaluation of the method and neural network for solving the problems of matching digital images was carried out. The neural network is implemented as a software utility Il.: 2; Bibliogr.: 15 titles.