Pairwise Feature Research Articles

Conditional Variational Autoencoder and generative adversarial network-based approach for long-tailed fault diagnosis for the motor system

This study addresses the challenge of diagnosing motor faults in long-tailed data distributions, characterized by dominant healthy states and rare fault types. We propose the LT-CVAE-GAN model, which integrates a Conditional Variational Autoencoder (CVAE) with a Conditional Generative Adversarial Network (CGAN) to enhance long-tailed fault diagnosis. Initially, we train the CVAE-GAN model using traditional CVAE and CGAN losses such as Kullback–Leibler (KL) divergence loss, reconstruction loss, and adversarial loss. Additionally, we introduce mean feature matching loss and pairwise feature matching loss to mitigate mode collapse and improve model stability, thereby enhancing the generation ability of less frequent fault samples under long-tail conditions. Subsequently, the pre-trained Generator is used to produce infrequent fault mode data to rebalance the dataset. Classifier parameters are fine-tuned in this step to improve fault diagnosis accuracy. Experimental results demonstrate that our LT-CVAE-GAN surpasses state-of-the-art models in diverse long-tailed conditions.

Scalable Methods for Multiple Time Series Comparison in Second Order Dynamics

Statistical comparison of multiple time series in their underlying frequency patterns has many real applications. However, existing methods are only applicable to a small number of mutually independent time series, and empirical results for dependent time series are only limited to comparing two time series. We propose scalable methods based on a new algorithm that enables us to compare the spectral density of a large number of time series. The new algorithm helps us efficiently obtain all pairwise feature differences in frequency patterns between M time series, which plays an essential role in our methods. When all M time series are independent of each other, we derive the joint asymptotic distribution of their pairwise feature differences. The asymptotic dependence structure between the feature differences motivates our proposed test for multiple mutually independent time series. We then adapt this test to the case of multiple dependent time series by partially accounting for the underlying dependence structure. Additionally, we introduce a global test to further enhance the approach. To examine the finite sample performance of our proposed methods, we conduct simulation studies. The new approaches demonstrate the ability to compare a large number of time series, whether independent or dependent, while exhibiting competitive power. Finally, we apply our methods to compare multiple mechanical vibrational time series.

Generative Adversarial Networks with Radiomics Supervision for Lung Lesion Generation.

Data-driven methods for lesion generation are quickly emerging due to the need for realistic imaging targets for image quality assessment and virtual clinical trials. We proposed a generative adversarial network (GAN) architecture for conditional generation of lung lesions based on user-specified classes of lesion size and solidity. The network consists of two discriminators, one for volumetric lesion data, and one for radiomics features derived from the lesion volume. A Wasserstein loss with gradient penalty was adopted for each discriminator. Training data were drawn from contoured and annotated lesions from a public lung CT database. Four quantitative evaluation methods were devised to assess the network performance: 1) overfitting (similarity between generated and real lesions), 2) diversity (similarity among generated lesions), 3) conditional consistency (capability of generating lesions according to user-specified classes), and 4) similarity in distributions of various lesion properties between the generated and real lesions. Ablation studies were also performed to investigate the importance of individual network component. The proposed network was found to generate lesions that resemble real lesions by visual inspection. Solid lesions are distinct from non-solid ones, and lesion sizes largely correspond to their specified classes. With a classifier trained on real lesions, the classification accuracies of generated and real lesions in both solid and non-solid classes are similar. Radiomics features of generated and real lesions were found to have similar distributions, indicated by the relatively low Kullback-Leibler (KL) divergence values. Furthermore, the correlations between pairwise radiomics features in generated lesions were comparable to those of real lesions. The proposed network presents a promising approach for generating realistic lesions with clinically relevant features crucial for the comprehensive assessment of medical imaging systems.

ARFIS: An adaptive robust model for regression with heavy-tailed distribution

As heavy-tailed distributions are ubiquitous in many real applications, robust regression has been extensively applied in machine learning and exhibits the superiority in deal with heavy-tailed distribution. Current existing robust methods for regression are based on independence assumption of features and ignore interaction between them, which may lead to poor generalization due to most datasets unsatisfying the assumption. Actually, interaction features formed by the composition of two or more features are particularly helpful to obtain the high-order information in many applications. In this paper, we propose a novel adaptive robust feature interaction selection model for regression with heavy-tailed distributions, termed Adaptive Robust Feature Interaction Selection (ARFIS). Firstly, we consider pairwise feature interaction by augmenting a feature vector with product of features for regression with heavy tailed distribution. Secondly, we propose feature interaction selection models based on quantile loss with different regularizers to learn parameters. The consistency of the proposed model ARFIS is theoretically proven, and an efficient algorithm is presented for solving proposed model. Finally, experimental results on simulation data, UCI datasets and a real-world dataset validate good accuracy, interpretability and robustness of our proposed models.

Machine-Learning-Assisted Design of Novel TiZrNbVAl Refractory High-Entropy Alloys with Enhanced Ductility

Refractory high-entropy alloys (RHEAs) typically exhibit excellent high-temperature strength but limited ductility. In this study, a comprehensive machine learning strategy with integrated material knowledge is proposed to predict the elongation of TiZrNbVAl RHEAs. By referring to the ductility theories, a set of cost-effective material features is developed with various mathematical forms of thermodynamic parameters. These features are proven to effectively incorporate material knowledge into ML modeling. They also offer potential alternatives to those obtained from costly first-principles calculations. Based on Pearson correlation coefficients, the linear relationships between pairwise features were compared, and the seven key features with the greatest impact on the model were selected for ML modeling. Regression tasks were performed to predict the ductility of TiZrNbVAl, and the CatBoost gradient boosting algorithm exhibiting the best performance was eventually selected. The established optimized model achieves high predictive accuracies exceeding 0.8. These key features were further analyzed using interpretable ML methods to elucidate their influences on various ductility mechanisms. According to the ML results, different compositions of TiZrNbVAl with excellent tensile properties were prepared. The experimental results indicate that Ti44Zr24Nb17V5Al10 and Ti44Zr26Nb8V13Al9 both exhibited ultimate tensile strengths of approximately 1180 MPa and elongations higher than 21%. They verified that the ML strategy proposed in this study is an effective approach for predicting the properties of RHEAs. It is a potential method that can replace costly first-principles calculations. Thermodynamic parameters have been shown to effectively predict alloy ductility to a certain extent.

PlaneSeg: Building a Plug-In for Boosting Planar Region Segmentation.

Existing methods in planar region segmentation suffer the problems of vague boundaries and failure to detect small-sized regions. To address these, this study presents an end-to-end framework, named PlaneSeg, which can be easily integrated into various plane segmentation models. Specifically, PlaneSeg contains three modules, namely, the edge feature extraction module, the multiscale module, and the resolution-adaptation module. First, the edge feature extraction module produces edge-aware feature maps for finer segmentation boundaries. The learned edge information acts as a constraint to mitigate inaccurate boundaries. Second, the multiscale module combines feature maps of different layers to harvest spatial and semantic information from planar objects. The multiformity of object information can help recognize small-sized objects to produce more accurate segmentation results. Third, the resolution-adaptation module fuses the feature maps produced by the two aforementioned modules. For this module, a pairwise feature fusion is adopted to resample the dropped pixels and extract more detailed features. Extensive experiments demonstrate that PlaneSeg outperforms other state-of-the-art approaches on three downstream tasks, including plane segmentation, 3-D plane reconstruction, and depth prediction. Code is available at https://github.com/nku-zhichengzhang/PlaneSeg.

Pairwise-Constraint-Guided Multi-View Feature Selection by Joint Sparse Regularization and Similarity Learning

Feature selection is a basic and important step in real applications, such as face recognition and image segmentation. In this paper, we propose a new weakly supervised multi-view feature selection method by utilizing pairwise constraints, i.e., the pairwise constraint-guided multi-view feature selection (PCFS for short) method. In this method, linear projections of all views and a consistent similarity graph with pairwise constraints are jointly optimized to learning discriminative projections. Meanwhile, the l2,0-norm-based row sparsity constraint is imposed on the concatenation of projections for discriminative feature selection. Then, an iterative algorithm with theoretically guaranteed convergence is developed for the optimization of PCFS. The performance of the proposed PCFS method was evaluated by comprehensive experiments on six benchmark datasets and applications on cancer clustering. The experimental results demonstrate that PCFS exhibited competitive performance in feature selection in comparison with related models.

Few-shot open-set recognition via pairwise discriminant aggregation

Few-shot open-set recognition (FSOR) aims to develop models capable of generalizing to new tasks for both unknown detection and known classification with limited labeled data. Previous FSOR methods lack a comprehensive definition of the open space in the few-shot scenario, making models susceptible to overgeneralization. In this paper, we propose a novel method called Pairwise Discriminant Aggregation (PDAgg), addressing FSOR within a two-level recognition framework. PDAgg unifies the diverse optimization goals of FSOR at the pair level and provides a reasonable aggregate-level representation for unknown samples, thereby greatly enhancing model generalization to open space in the few-shot context. Specifically, PDAgg treats support-query pairs as the basic recognition units, which are adapted to a pair-specific feature space by enhancing pairwise representative features and incorporating a global knowledge context. Binary discriminant analyses are performed on adapted pair embeddings to estimate pair-level discriminant scores, which are then jointly aggregated to achieve both unknown detection and few-shot classification. Extensive experiments demonstrate that our method delivers comparable and even better performance with less extra information compared to the existing FSOR methods.

Psychology-Guided Environment Aware Network for Discovering Social Interaction Groups from Videos

Social interaction is a common phenomenon in human societies. Different from discovering groups based on the similarity of individuals’ actions, social interaction focuses more on the mutual influence between people. Although people can easily judge whether or not there are social interactions in a real-world scene, it is difficult for an intelligent system to discover social interactions. Initiating and concluding social interactions are greatly influenced by an individual’s social cognition and the surrounding environment, which are closely related to psychology. Thus, converting the psychological factors that impact social interactions into quantifiable visual representations and creating a model for interaction relationships poses a significant challenge. To this end, we propose a Psychology-Guided Environment Aware Network (PEAN) that models social interaction among people in videos using supervised learning. Specifically, we divide the surrounding environment into scene-aware visual-based and human-aware visual-based descriptions. For the scene-aware visual clue, we utilize 3D features as global visual representations. For the human-aware visual clue, we consider instance-based location and behaviour-related visual representations to map human-centred interaction elements in social psychology: distance, openness, and orientation. In addition, we design an environment aware mechanism to integrate features from visual clues, with a Transformer to explore the relation between individuals and construct pairwise interaction strength features. The interaction intensity matrix reflecting the mutual nature of the interaction is obtained by processing the interaction strength features with the interaction discovery module. An interaction constrained loss function composed of interaction critical loss function and smooth F β loss function is proposed to optimize the whole framework to improve the distinction of the interaction matrix and alleviate class imbalance caused by pairwise interaction sparsity. Given the diversity of real-world interactions, we collect a new dataset named Social Basketball Activity Dataset (Soical-BAD), covering complex social interactions. Our method achieves the best performance among social-CAD, social-BAD, and their combined dataset named Video Social Interaction Dataset (VSID).

Assessment and classification of COVID-19 DNA sequence using pairwise features concatenation from multi-transformer and deep features with machine learning models

The 2019 novel coronavirus (renamed SARS-CoV-2, and generally referred to as the COVID-19 virus) has spread to 184 countries with over 1.5 million confirmed cases. Such a major viral outbreak demands early elucidation of taxonomic classification and origin of the virus genomic sequence, for strategic planning, containment, and treatment. The emerging global infectious COVID-19 disease by novel Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) presents critical threats to global public health and the economy since it was identified in late December 2019 in China. The virus has gone through various pathways of evolution. Due to the continued evolution of the SARS-CoV-2 pandemic, researchers worldwide are working to mitigate, suppress its spread, and better understand it by deploying deep learning and machine learning approaches. In a general computational context for biomedical data analysis, DNA sequence classification is a crucial challenge. Several machine and deep learning techniques have been used in recent years to complete this task with some success. The classification of DNA sequences is a key research area in bioinformatics as it enables researchers to conduct genomic analysis and detect possible diseases. In this paper, three state-of-the-art deep learning-based models are proposed using two DNA sequence conversion methods. We also proposed a novel multi-transformer deep learning model and pairwise features fusion technique for DNA sequence classification. Furthermore, deep features are extracted from the last layer of the multi-transformer and used in machine-learning models for DNA sequence classification. The k-mer and one-hot encoding sequence conversion techniques have been presented. The proposed multi-transformer achieved the highest performance in COVID DNA sequence classification. Automatic identification and classification of viruses are essential to avoid an outbreak like COVID-19. It also helps in detecting the effect of viruses and drug design.

Cognitive pairwise comparison forward feature selection with deep learning for astronomical object classification with sloan digital sky survey

This paper proposes a hybrid approach integrating the expert knowledge judgment approach using the Cognitive Pairwise Comparison (CPC) to the Deep Learning, a modern classification approach, for astronomic object classification. The astronomic data with ten thousand samples retrieved from Sloan Digital Sky Survey Sky Server Data Release 15 (SDSS SkyServer DR 15) are used for this study. The CPC is an approach to elicit and encode expert knowledge in the format of a Pairwise Opposite Matrix (POM) to evaluate expert preferences for the features. A forward feature selection algorithm taking the expert choices using CPC for the ordered features is used for the feature selection for the deep learning algorithm to build a heuristic training model based on the astronomic data. Whilst the accuracy of the case of improper feature selection is just 37.1%, the proposed hybrid approach can obtain a very high accuracy of 97.9% for the classification of the astronomic object using the eight scaled features (u, g, r, i, z redshift, ra, dec). To extend this research, the proposed CPC can be used as a human-centered tool to be applied to other areas of data sciences.

Automated generative design and prefabrication of precast buildings using integrated BIM and graph convolutional neural network

Precast construction is a productivity-improving technology in the architectural, engineering, and construction industry that improves construction efficiency by combining factory-based manufacturing and lean assembly. Many international efforts have encouraged the adoption of this approach. This study presents an integrated Building Information Modelling (BIM) with technological automation interoperability to enable generative design and prefabrication for precast buildings. A generic BIM-based graph representation is established to explicitly formulate buildings' spatial and geometric features. Following this, a graph-constrained layout generator is developed, with a generative modelling algorithm and graph convolutional neural network, to extract pairwise spatial-geometric features for generating the optimal precast layout. This is followed by semantic enrichment of BIM data (i.e., Industry Foundation Classes) with precast data schema to facilitate data transformation for prefabrication automation until site delivery. The holistic approach presented in this study empowers pre-construction planning optimisation and fabrication automation in precast construction.

Learning Association Characteristics by Dynamic Hypergraph and Gated Convolution Enhanced Pairwise Attributes for Prediction of Disease-Related lncRNAs.

As the long non-coding RNAs (lncRNAs) play important roles during the incurrence and development of various human diseases, identifying disease-related lncRNAs can contribute to clarifying the pathogenesis of diseases. Most of the recent lncRNA-disease association prediction methods utilized the multi-source data about the lncRNAs and diseases. A single lncRNA may participate in multiple disease processes, and multiple lncRNAs usually are involved in the same disease process synergistically. However, the previous methods did not completely exploit the biological characteristics to construct the informative prediction models. We construct a prediction model based on adaptive hypergraph and gated convolution for lncRNA-disease association prediction (AGLDA), to embed and encode the biological characteristics about lncRNA-disease associations, the topological features from the entire heterogeneous graph perspective, and the gated enhanced pairwise features. First, the strategy for constructing hyperedges is designed to reflect the biological characteristic that multiple lncRNAs are involved in multiple disease processes. Furthermore, each hyperedge has its own biological perspective, and multiple hyperedges are beneficial for revealing the diverse relationships among multiple lncRNAs and diseases. Second, we encode the biological features of each lncRNA (disease) node using a strategy based on dynamic hypergraph convolutional networks. The strategy may adaptively learn the features of the hyperedges and formulate the dynamically evolved hypergraph topological structure. Third, a group convolutional network is established to integrate the entire heterogeneous topological structure and multiple types of node attributes within an lncRNA-disease-miRNA graph. Finally, a gated convolutional strategy is proposed to enhance the informative features of the lncRNA-disease node pairs. The comparison experiments indicate that AGLDA outperforms seven advanced prediction methods. The ablation studies confirm the effectiveness of major innovations, and the case studies validate AGLDA's ability in application for discovering potential disease-related lncRNA candidates.

Structure shapes the representation of a novel category.

Concepts contain rich structures that support flexible semantic cognition. These structures can be characterized by patterns of feature covariation: Certain features tend to cluster in the same items (e.g., feathers, wings, can fly). Existing computational models demonstrate how this kind of structure can be leveraged to slowly learn the distinctions between categories, on developmental timescales. However, it is not clear whether and how we leverage feature structure to quickly learn a novel category. We thus investigated how the internal structure of a new category is first extracted from experience, with the prediction that feature-based structure would have a rapid and broad influence on the learned category representation. Across three experiments, novel categories were designed with patterns of feature associations determined by carefully constructed graph structures, with Modular graphs-exhibiting strong clusters of feature covariation-compared against Random and Lattice graphs. In Experiment 1, a feature inference task using verbal stimuli revealed that Modular structure broadly facilitated category learning. Experiment 2 replicated this effect in visual categories. In Experiment 3, a statistical learning paradigm revealed that this Modular benefit relates to high-level structure rather than pairwise feature associations and persists even when category structure is incidental to the task. A neural network model was readily able to account for these effects, suggesting that correlational feature structure may be encoded within rapidly learned, distributed category representations. These findings constrain theories of category representation and link theories of category learning with structure learning more broadly. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Pairwise CNN-Transformer Features for Human-Object Interaction Detection.

Human-object interaction (HOI) detection aims to localize and recognize the relationship between humans and objects, which helps computers understand high-level semantics. In HOI detection, two-stage and one-stage methods have distinct advantages and disadvantages. The two-stage methods can obtain high-quality human-object pair features based on object detection but lack contextual information. The one-stage transformer-based methods can model good global features but cannot benefit from object detection. The ideal model should have the advantages of both methods. Therefore, we propose the Pairwise Convolutional neural network (CNN)-Transformer (PCT), a simple and effective two-stage method. The model both fully utilizes the object detector and has rich contextual information. Specifically, we obtain pairwise CNN features from the CNN backbone. These features are fused with pairwise transformer features to enhance the pairwise representations. The enhanced representations are superior to using CNN and transformer features individually. In addition, the global features of the transformer provide valuable contextual cues. We fairly compare the performance of pairwise CNN and pairwise transformer features in HOI detection. The experimental results show that the previously neglected CNN features still have a significant edge. Compared to state-of-the-art methods, our model achieves competitive results on the HICO-DET and V-COCO datasets.

Relation fusion propagation network for transductive few-shot learning

Previous graph-based meta-learning approaches have explored pairwise feature similarity to learn instance-level relations of samples, however, the gap between the sample relations in feature and label spaces is often ignored. It is empirically observed that instances with different labels may display considerable similarity in visual characteristics, making it challenging to distinguish between them in the feature space. To this end, we propose a dual-branch Relation Fusion Propagation Network (RFPN) for transductive few-shot learning, which explicitly models both feature and label relations across support-query pairs. Specifically, we design a Relation Fusion Block (RFB) to fuse instance-level and class-level relations, thus obtaining more robust fusion-level relations to guide feature propagation. In addition to the feature propagation branch, we encode the label relations and construct a label shortcut branch for label propagation. To alleviate the error accumulation during propagation, which is caused by uncertain pseudo-labels for query samples, we propose a Label Shortcut Mechanism (LSM) to progressively update the sample relations with the initial labels. Our full method is plug-and-play and can be easily applied in existing graph-based approaches for transductive few-shot learning. Extensive experiments demonstrate that our proposed RFPN yields significant improvements over the baselines and achieves promising performance on four popular few-shot classification benchmarks.

FET-FGVC: Feature-enhanced transformer for fine-grained visual classification

The challenge of Fine-grained visual classification (FGVC) comes from the small variations between classes and the large variations within classes. Inspired by the fact that identifying bird species focuses not only on the global features of the subject area but also on the subtle details of the local area, we propose a feature-enhanced Transformer to improve the performance of FGVC. Our proposed method consists of a Dynamic Swin Transformer backbone for extracting comprehensive global image features through continuous attention aggregation, a GCN-based local branch for separating and enhancing local features in different regions, and a pairwise feature interaction (PFI) module for enhancing global features through interactions between image pairs. We conducted extensive experiments on five FGVC datasets to demonstrate the superiority of our method. By fusing the enhanced global and local features, our method achieves the best accuracy compared to existing methods. Our method has an advantage in terms of computational efficiency.

Skin Conductance-Based Acupoint and Non-Acupoint Recognition Using Machine Learning.

Acupoints (APs) prove to have positive effects on disease diagnosis and treatment, while intelligent techniques for the automatic detection of APs are not yet mature, making them more dependent on manual positioning. In this paper, we realize the skin conductance-based APs and non-APs recognition with machine learning, which could assist in APs detection and localization in clinical practice. Firstly, we collect skin conductance of traditional Five-Shu Point and their corresponding non-APs with wearable sensors, establishing a dataset containing over 36000 samples of 12 different AP types. Then, electrical features are extracted from the time domain, frequency domain, and nonlinear perspective respectively, following which typical machine learning algorithms (SVM, RF, KNN, NB, and XGBoost) are demonstrated to recognize APs and non-APs. The results demonstrate XGBoost with the best precision of 66.38%. Moreover, we also quantify the impacts of the differences among AP types and individuals, and propose a pairwise feature generation method to weaken the impacts on recognition precision. By using generated pairwise features, the recognition precision could be improved by 7.17%. The research systematically realizes the automatic recognition of APs and non-APs, and is conducive to pushing forward the intelligent development of APs and Traditional Chinese Medicine theories.

GSSF: Generalized Structural Sparse Function for Deep Cross-Modal Metric Learning.

Cross-modal metric learning is a prominent research topic that bridges the semantic heterogeneity between vision and language. Existing methods frequently utilize simple cosine or complex distance metrics to transform the pairwise features into a similarity score, which suffers from an inadequate or inefficient capability for distance measurements. Consequently, we propose a Generalized Structural Sparse Function to dynamically capture thorough and powerful relationships across modalities for pair-wise similarity learning while remaining concise but efficient. Specifically, the distance metric delicately encapsulates two formats of diagonal and block-diagonal terms, automatically distinguishing and highlighting the cross-channel relevancy and dependency inside a structured and organized topology. Hence, it thereby empowers itself to adapt to the optimal matching patterns between the paired features and reaches a sweet spot between model complexity and capability. Extensive experiments on cross-modal and two extra uni-modal retrieval tasks (image-text retrieval, person re-identification, fine-grained image retrieval) have validated its superiority and flexibility over various popular retrieval frameworks. More importantly, we further discover that it can be seamlessly incorporated into multiple application scenarios, and demonstrates promising prospects from Attention Mechanism to Knowledge Distillation in a plug-and-play manner.

An automated ECG-based deep learning for the early-stage identification and classification of cardiovascular disease.

Heart disease represents the leading cause of death globally. Timely diagnosis and treatment can prevent cardiovascular issues. An Electrocardiograms (ECG) serves as a diagnostic tool for identifying heart difficulties. Cardiovascular Disease (CVD) often gets identified through ECGs. Deep learning (DL) garners attention in healthcare due to its potential in swiftly diagnosing ECG anomalies, crucial for patient monitoring. Conversely, automatic CVD detection from ECGs poses a challenging task, wherein rule-based diagnostic models usually achieve top-notch performance. These models encounter complications in supervision vast volumes of diverse data, demanding widespread analysis and medical capability to ensure precise CVD diagnosis. This study aims to enhance cardiovascular disease diagnosis by combining symptom-based detection and ECG analysis. To enhance these experiments, we built a novel automated prediction method based on a Feed Forward Neural Network (FFNN) model. The fundamental objective of our method is to develop the accuracy of ECG diagnosis. Our strategy employs chaos theory and destruction analysis to combine optimum deep learning features with a well-organized set of ECG properties. In addition, we use the constant-Q non-stationary Gabor transform (CQNGT) to convert one-dimensional ECG data into a two-dimensional picture. A pre-trained FFNN processes this image. To identify significant features from the FFNN output that correspond with the ECG data, we employ pairwise feature proximity. According to experimental findings, the suggested system, FFNN-CQNGT, surpasses other state-of-the-art systems in terms of precision of 94.89%, computational efficiency of 2.114 ms, accuracy of 95.55%, specificity of 93.77%, and sensitivity of 93.99% and MSE 40.32%. Contributing an automated ECG-based DL system based on FFNN-CQNGT for early-stage cardiovascular disease identification and classification holds great potential for both patient care and public health.