Structural Information Research Articles

JTIS: enhancing biomedical document-level relation extraction through joint training with intermediate steps

Abstract Biomedical Relation Extraction (RE) is central to Biomedical Natural Language Processing and is crucial for various downstream applications. Existing RE challenges in the field of biology have primarily focused on intra-sentential analysis. However, with the rapid increase in the volume of literature and the complexity of relationships between biomedical entities, it often becomes necessary to consider multiple sentences to fully extract the relationship between a pair of entities. Current methods often fail to fully capture the complex semantic structures of information in documents, thereby affecting extraction accuracy. Therefore, unlike traditional RE methods that rely on sentence-level analysis and heuristic rules, our method focuses on extracting entity relationships from biomedical literature titles and abstracts and classifying relations that are novel findings. In our method, a multitask training approach is employed for fine-tuning a Pre-trained Language Model in the field of biology. Based on a broad spectrum of carefully designed tasks, our multitask method not only extracts relations of better quality due to more effective supervision but also achieves a more accurate classification of whether the entity pairs are novel findings. Moreover, by applying a model ensemble method, we further enhance our model’s performance. The extensive experiments demonstrate that our method achieves significant performance improvements, i.e. surpassing the existing baseline by 3.94% in RE and 3.27% in Triplet Novel Typing in F1 score on BioRED, confirming its effectiveness in handling complex biomedical literature RE tasks. Database URL: https://codalab.lisn.upsaclay.fr/competitions/13377#learn_the_details-dataset

Isomer Differentiation by UHPLC-Q-Exactive-Orbitrap MS led to Enhanced Identification of Daphnane Diterpenoids in Daphne tangutica.

Liquid chromatography-mass spectrometry (LC-MS) has enhanced the rapid, accurate analysis of complex plant extracts, eliminating the need for extensive isolation. Tandem mass spectrometry (MS/MS) further enhances this process by providing detailed structural information. However, differentiating structural isomers remains a challenge due to their minor spectral and structural differences. This study aimed to extend the applicability of LC-MS/MS for the structural identification of daphnane diterpenoids, with a particular focus on distinguishing functional isomers. LC-MS analyses were performed using an UHPLC-Q-Exactive-Orbitrap MS. The MS conditions for distinguishing isomers were optimized using in-source CID and HCD modes with reference compounds. A qualitative analysis was then conducted on the extract of Daphne tangutica. The chemical structures of the detected daphnane diterpenoids were estimated by analyzing the fragmentation patterns in both the mass spectra and product ion spectra. These identifications were further validated by isolation and comparison with an in-house daphnane diterpenoid library. By optimizing MS conditions, especially in the negative ion mode, it was possible to accurately distinguish structural isomers such as yuanhuajine and gniditrin. Qualitative analysis of D. tangutica identified a total of 28 daphnanes, including seven previously unreported compounds. Furthermore, a novel geometric isomer of gniditrin was isolated by conducting isolation on the crude diterpenoid fraction. This study demonstrated that LC-MS/MS analysis can effectively distinguish functional isomers of daphnane diterpenoids, thereby enhancing the identification of daphnanes in plant extracts and highlighting its potential as a powerful tool for phytochemical analysis.

A coarse-to-fine image registration method based on autocorrelation structural difference information

ABSTRACT The automatic registration of synthetic aperture radar (SAR) and optical images is still a challenging problem due to different imaging mechanisms. This letter proposes a coarse-to-fine image registration method that leverages self-similarity structural difference information. In the coarse registration stage, a Scale-Invariant Feature Transformation-based (SIFT-like) method is employed, complemented by an improved Fast Sample Consensus (IFSC) method to eliminate mismatched point pairs by probabilistic and geometric information. This stage ensures robustness against scale and rotational variations. In the fine registration stage, robust feature points are selected by utilizing phase and edge structural information. A descriptor which based on phase consistency and autocorrelation structural difference (ASDPC) is constructed to capture the structural variations between region blocks, and a fine search is carried out within the neighbourhood of the already matched feature points, so as to find more accurate matched feature points and obtain fine registration. The experimental results demonstrate that the proposed method provides robust and accurate registration for optical-to-SAR images.

Adaptive Centroid-Connected Structure Matching Network Based on Semi-Supervised Heterogeneous Domain

Heterogeneous domain adaptation (HDA) utilizes the knowledge of the source domain to model the target domain. Although the two domains are semantically related, the problem of feature and distribution differences in heterogeneous data still needs to be solved. Most of the existing HDA methods only consider the feature or distribution problem but do not consider the geometric semantic information similarity between the domain structures, which leads to a weakened adaptive performance. In order to solve the problem, a centroid connected structure matching network (CCSMN) approach is proposed, which firstly maps the heterogeneous data into a shared public feature subspace to solve the problem of feature differences. Secondly, it promotes the overlap of domain centers and nodes of the same category between domains to reduce the positional distribution differences in the internal structure of data. Then, the supervised information is utilized to generate target domain nodes, and the geometric structural and semantic information are utilized to construct a centroid-connected structure with a reasonable inter-class distance. During the training process, a progressive and integrated pseudo-labeling is utilized to select samples with high-confidence labels and improve the classification accuracy for the target domain. Extensive experiments are conducted in text-to-image and image-to-image HDA tasks, and the results show that the CCSMN outperforms several state-of-the-art baseline methods. Compared with state-of-the-art HDA methods, in the text-to-image transfer task, the efficiency has increased by 8.05%; and in the image-to-image transfer task, the efficiency has increased by about 2%, which suggests that the CCSMN benefits more from domain geometric semantic information similarity.

Structural variant allelic heterogeneity in MECP2 duplication syndrome provides insight into clinical severity and variability of disease expression

BackgroundMECP2 Duplication Syndrome, also known as X-linked intellectual developmental disorder Lubs type (MRXSL; MIM: 300260), is a neurodevelopmental disorder caused by copy number gains spanning MECP2. Despite varying genomic rearrangement structures, including duplications and triplications, and a wide range of duplication sizes, no clear correlation exists between DNA rearrangement and clinical features. We had previously demonstrated that up to 38% of MRXSL families are characterized by complex genomic rearrangements (CGRs) of intermediate complexity (2 ≤ copy number variant breakpoints < 5), yet the impact of these genomic structures on regulation of gene expression and phenotypic manifestations have not been investigated.MethodsTo study the role of the genomic rearrangement structures on an individual’s clinical phenotypic variability, we employed a comprehensive genomics, transcriptomics, and deep phenotyping analysis approach on 137 individuals affected by MRXSL. Genomic structural information was correlated with transcriptomic and quantitative phenotypic analysis using Human Phenotype Ontology (HPO) semantic similarity scores.ResultsDuplication sizes in the cohort ranging from 64.6 kb to 16.5 Mb were classified into four categories comprising of tandem duplications (48%), terminal duplications (22%), inverted triplications (20%), and other CGRs (10%). Most of the terminal duplication structures consist of translocations (65%) followed by recombinant chromosomes (23%). Notably, 65% of de novo events occurred in the Terminal duplication group in contrast with 17% observed in Tandem duplications. RNA-seq data from lymphoblastoid cell lines indicated that the MECP2 transcript quantity in MECP2 triplications is statistically different from all duplications, but not between other classes of genomic structures. We also observed a significant (p < 0.05) correlation (Pearson R = 0.6, Spearman p = 0.63) between the log-transformed MECP2 RNA levels and MECP2 protein levels, demonstrating that genomic aberrations spanning MECP2 lead to altered MECP2 RNA and MECP2 protein levels. Genotype–phenotype analyses indicated a gradual worsening of phenotypic features, including overall survival, developmental levels, microcephaly, epilepsy, and genitourinary/eye abnormalities in the following order: Tandem duplications, Other complex duplications, Terminal duplications/Translocations, and Triplications encompassing MECP2.ConclusionIn aggregate, this combined analysis uncovers an interplay between MECP2 dosage, genomic rearrangement structure and phenotypic traits. Whereas the level of MECP2 is a key determinant of the phenotype, the DNA rearrangement structure can contribute to clinical severity and disease expression variability. Employing this type of analytical approach will advance our understanding of the impact of genomic rearrangements on genomic disorders and may help guide more targeted therapeutic approaches.

Molecular identification, genetic diversity, and secondary structure predictions of Physalis species using ITS2 DNA barcoding

BackgroundThe genus Physalis belongs to the Solanaceae family and has different species with important nutritional and medicinal values. Species within this genus have limited morphological differences, a characteristic that hinders accurate identification, safe utilization and genetic conservation of promising genotypes. In addition, to prevent the perceived loss of Physalis diversity due to habitat destruction, species delimitation needs attention. In this study, we used the sequence and structural information of the internal transcribed spacer 2 (ITS2) barcode to efficiently identify and discriminate Physalis species from a collection of 34 Physalis accessions.MethodologyPhysalis plant samples were collected from eight Counties in Kenya based on the availability of the germplasm. The voucher specimens were identified using the botanical taxonomy method and were deposited in the University of Nairobi herbarium. A total of 34 Physalis accessions were identified and accessed for diversity based on the ITS2 barcode region. The sequence similarity of the ITS2 genes was analyzed through the Basic Local Alignment Search Tool (BLAST), the nearest Kimura-2-parameter (K2P) genetic distances were calculated and a phylogenetic tree was constructed using the Bayesian inference (BI) method in MrBayes 3.2.7a software. The differences in the ITS2 secondary structure between the species were analyzed.ResultsThe success rate of PCR amplification and sequencing was 75% and 67%, respectively. The analyzed ITS2 sequences displayed significant inter-specific divergences, clear DNA barcoding gaps and high species identification efficiency. Based on the constructed phylogenetic tree, three Physalis species (Physalis peruviana, Physalis purpurea and Physalis cordata) were identified and were clustered in a homogenized distribution. High genetic diversity (0.36923) and genetic distance (0.703) were observed between Physalis peruviana and Physalis cordata. The highest genetic nucleotide diversity (0.26324) and distance (0.46) within species was obtained for Physalis peruviana. The differences in the secondary structures generated from this study discriminated between the Physalis species.ConclusionsOur study demonstrated that ITS2 is a potential DNA barcode for effective identification and discrimination of Physalis species. The results of this study provide insights into the scientific basis of species identification, safe utilization, genetic conservation and future breeding strategies for this important nutritional and medicinal plant species.

Computer-assisted medical history taking prior to patient consultation in the outpatient care setting: a prospective pilot project

BackgroundFeeding patients’ self-reported medical history into the diagnostic care process may accelerate workflows in clinical routine.MethodsWe prospectively piloted a novel medical history documentation system in a German cardiological outpatient practice and evaluated its feasibility and perceived usefulness. Based on a generic software that allows to record structured information, a customized solution for the cooperating practice was developed and implemented. Prior to the consultation of the physician, the patient used a tablet that guided the user through a structured comprehensive workflow to document the medical history. The retrieved information was arranged by the software into a ready-to-use text format, presented to the physician in an editable form and added to her report. Three user-centered endpoints were explored: i) Appropriateness—measured by the duration of a patient interview; ii) Patient acceptance—assessed by three questions to patients; iii) Usefulness—operationalized by multiple ratings of the physician.ResultsA total of 2,513 patients were approached of which 2,415 provided complete histories. The system was assessed as appropriate for the practical workflow in terms of time and workflows. The patient-system interaction was rated favourably by patients including elderly ones. The system was regarded useful by the physician, reducing her daily workload by about one hour.ConclusionsAutomated history-taking tools deployed before consultation could support physicians in obtaining patients’ medical histories, thereby reducing professionals’ perceived workload. The technical and methodological limitations of our study should be respected, calling for additional future evaluations.

DiffInt: A Diffusion Model for Structure-Based Drug Design with Explicit Hydrogen Bond Interaction Guidance.

The design of drug molecules is a critical stage in the drug discovery process. The structure-based drug design has long played an important role in efficient development. Significant progress has been made in recent years in the generation of 3D molecules via deep generation models. However, while many existing models have succeeded in incorporating structural information on target proteins, they have not been able to address important interactions between protein and drug molecules, especially hydrogen bonds. In this study, we propose DiffInt as a novel structure-based approach that explicitly addresses interactions. The model naturally incorporates hydrogen bonds between protein and ligand molecules by treating them as pseudoparticles. The experimental results show that DiffInt reproduces hydrogen bonds, and the hydrogen binding energies significantly outperform those of existing models. To facilitate the use of our tool for generating new drug molecules based on any protein's three-dimensional structure, we have made the source code and trained model available on GitHub (https://github.com/sekijima-lab/DiffInt) under the MIT license, with the execution environment provided on Google Colab.

Older Adults’ Disorientation in Smartphones: Exploring the Role of Spatial Ability and Quantifying Cognitive Similarity Between Perceived and Physical Information Structure

Older adults are prone to getting lost when switching between screens on smartphones. Navigating through screens can be demanding regarding the user’s spatial ability, which tends to decline with age. This study investigates the impact of information structure and spatial ability on older adults’ navigation in smartphone applications. It proposes a method of quantifying cognitive similarity between older adults’ perceived information structure and designers’ information structure used in applications. Researchers designed three different structured applications as prototypes, with low, medium, and high levels of netting, but with the same contents. Thirty older adults over 65 years of age and thirty younger adults were recruited to participate in the experiment, completing two types of navigation tasks. There were three main findings from our experiment. Firstly, the more reticulated the information structure, the longer the navigation task took the older adults to complete and the more errors were returned, which was the opposite of the results for younger adults. Second, spatial ability has no significant impact on the performance of the process page navigation task. However, it has a significant correlation with the cognitive similarity of older adults to a certain information structure. The higher the score of spatial memory span and pointing ability, the more accurate the cognition construction. Thirdly, the fitting degree of user mind and information structure can be predicted on the premise of specific information structure and given user spatial ability.

Lighting the Path: Raman Spectroscopy’s Journey Through the Microbial Maze

The rapid and precise identification of microorganisms is essential in environmental science, pharmaceuticals, food safety, and medical diagnostics. Raman spectroscopy, valued for its ability to provide detailed chemical and structural information, has gained significant traction in these fields, especially with the adoption of various excitation wavelengths and tailored optical setups. The choice of wavelength and setup in Raman spectroscopy is influenced by factors such as applicability, cost, and whether bulk or single-cell analysis is performed, each impacting sensitivity and specificity in bacterial detection. In this study, we investigate the potential of different excitation wavelengths for bacterial identification, utilizing a mock culture composed of six bacterial species: three Gram-positive (S. warneri, S. cohnii, and E. malodoratus) and three Gram-negative (P. stutzeri, K. terrigena, and E. coli). To improve bacterial classification, we applied machine learning models to analyze and extract unique spectral features from Raman data. The results indicate that the choice of excitation wavelength significantly influences the bacterial spectra obtained, thereby impacting the accuracy and effectiveness of the subsequent classification results.

Adltformer Team-Training with Detr: Enhancing Cattle Detection in Non-Ideal Lighting Conditions Through Adaptive Image Enhancement

This study proposes an image enhancement detection technique based on Adltformer (Adaptive dynamic learning transformer) team-training with Detr (Detection transformer) to improve model accuracy in suboptimal conditions, addressing the challenge of detecting cattle in real pastures under complex lighting conditions—including backlighting, non-uniform lighting, and low light. This often results in the loss of image details and structural information, color distortion, and noise artifacts, thereby compromising the visual quality of captured images and reducing model accuracy. To train the Adltformer enhancement model, the day-to-night image synthesis (DTN-Synthesis) algorithm generates low-light image pairs that are precisely aligned with normal light images and include controlled noise levels. The Adltformer and Detr team-training (AT-Detr) method is employed to preprocess the low-light cattle dataset for image enhancement, ensuring that the enhanced images are more compatible with the requirements of machine vision systems. The experimental results demonstrate that the AT-Detr algorithm achieves superior detection accuracy, with comparable runtime and model complexity, reaching 97.5% accuracy under challenging illumination conditions, outperforming both Detr alone and sequential image enhancement followed by Detr. This approach provides both theoretical justification and practical applicability for detecting cattle under challenging conditions in real-world farming environments.

Multitiered computational methodology for extracting three-dimensional rotational diffusion coefficients from x-ray photon correlation spectroscopy data without structural information

Multitiered computational methodology for extracting three-dimensional rotational diffusion coefficients from x-ray photon correlation spectroscopy data without structural information

Residue-Level Multiview Deep Learning for ATP Binding Site Prediction and Applications in Kinase Inhibitors.

Accurate identification of adenosine triphosphate (ATP) binding sites is crucial for understanding cellular functions and advancing drug discovery, particularly in targeting kinases for cancer treatment. Existing methods face significant challenges due to their reliance on time-consuming precomputed features and the heavily imbalanced nature of binding site data without further investigations on their utility in drug discovery. To address these limitations, we introduced Multiview-ATPBind and ResiBoost. Multiview-ATPBind is an end-to-end deep learning model that integrates one-dimensional (1D) sequence and three-dimensional (3D) structural information for rapid and precise residue-level pocket-ligand interaction predictions. Additionally, ResiBoost is a novel residue-level boosting algorithm designed to mitigate data imbalance by enhancing the prediction of rare positive binding residues. Our approach outperforms state-of-the-art models on benchmark data sets, showing significant improvements in balanced metrics with both experimental and AI-predicted structures. Furthermore, our model seamlessly transfers to predicting binding sites and enhancing docking simulations for kinase inhibitors, including imatinib and dasatinib, underscoring the potential of our method in drug discovery applications.

Protein Modeling with DEER Spectroscopy.

Double electron-electron resonance (DEER) combined with site-directed spin labeling can provide distance distributions between selected protein residues to investigate protein structure and conformational heterogeneity. The utilization of the full quantitative information contained in DEER data requires effective protein and spin label modeling methods. Here, we review the application of DEER data to protein modeling. First, we discuss the significance of spin label modeling for accurate extraction of protein structural information and review the most popular label modeling methods. Next, we review several important aspects of protein modeling with DEER, including site selection, how DEER restraints are applied, common artifacts, and the unique potential of DEER data for modeling structural ensembles and conformational landscapes. Finally, we discuss common applications of protein modeling with DEER data and provide an outlook.

Solute Energetics in Aqueous Xanthan Gum Solutions: What Can Be Learned from a Fluorescent Probe?

Xanthan gum (XG) is a well-known anionic polysaccharide that finds broad application in the food and petroleum industries because of its ability to enhance solution viscosity at low concentrations and moderate temperatures. The aim of this work was to use the solvation probe coumarin 153 (C153) to characterize changes in the xanthan gum (XG) solution microstructure as a function of XG concentration and temperature from the perspective of a dissolved solute molecule. We established the utility of C153 fluorescence to track solution changes for XG concentrations that span the transition region from a dilute to a semi-dilute solution, defined by the xanthan gum overlap concentration, C*~0.02 g/dL. The temperature was varied from 293 to 353 K to probe solution conditions wherein XG has been reported to undergo a structural change from helix to random coil conformation, the details of which are still under debate. While C153 fluorescence does not elucidate direct structural information, the emission response is a simple means by which changes in aqueous XG solution can be identified. C153 spectroscopy is observed to correlate with XG conformational changes, as reported in the literature.

Machine Learning Application for High-Speed FTIR Absorption Spectra Analysis

"With this work our main objective is to find the best general (baseline) model for analyzing unknown spectra using Fourier infrared transformed spectroscopy (FT-IR) coupled with machine learning (ML) algorithms. This goal allows us to identify the best methodology applied for inline analysis of different experimental spectra for qualitative structural information obtained with types of structures that generate absorption or emission peaks. This methodology opens new perspectives for automated data processing using flexible algorithms and machine learning to encode experimental data for future applications. The results provide a good perspective on Machine Learning algorithms for applied sciences research. For our case study (FTIR experimental data) our model allows for peak feature extraction with a relative low, close to machine standard deviation, error budget. The best model identified is a specialized model but the standard, fully connected network models are evaluated."

Yingde–Guangning Granitic Plutons Record Complex Tectonic Evolution During Paleozoic-Mesozoic: Implications for Gold Exploration in Western Guangdong, South China

Western Guangdong, a part of the South China Block, has a complex geological history characterized by significant magmatic, metamorphic, and tectonic activities. This dynamic geological past, particularly during the Mesozoic era, created favorable conditions for the formation of various mineral deposits, including Au, Ag, Cu, and Pb. This makes the region a key area for precious metal resources in China. Despite extensive metallogenic studies, detailed structural information for western Guangdong remains insufficient, highlighting the need for further investigation. Thus, effective delineation of deformation periods is crucial for revealing geodynamic history and understanding regional tectonic activities, which are extremely important for guiding mineral exploration. This work focuses on the outcrops of granitic plutons in the Yingde–Guangning area of western Guangdong to establish the structure–tectonic setting. The tectonic events likely shaped the widespread Paleozoic–Mesozoic granitic bodies, which record extensive information on regional tectonic evolution. To achieve the primary objective, systematic identification and kinematic analysis of the various stages of structural traces, such as foliations and joints, have been conducted. This research proposes, for the first time, that the western Guangdong area underwent four distinct tectonic stages: (1) Early Paleozoic NW-SE compression phase; (2) Triassic NE-SW compressional stress; (3) Jurassic NW-SE compressional force; and (4) Cretaceous NW-SE extension stage. In metallogenic terms, the NW-SE trending auriferous veins of the Yingde–Guangning region were mostly formed during the Triassic NE-SW compression stage, whereas the NE-SW trending vein-type gold mineralization developed during the tectonic regime transformation from Jurassic NW-SE compression to Cretaceous NW-SE extension. This research emphasizes that systematic tectonic geological studies of regional granites can effectively guide mineral prospecting.

Metadata Forensic Analysis as Support for Digital Investigation Process by Utilizing Metadata-Extractor

Abstract — The rapid development of technology in the current era, in addition to providing positive impacts, certainly also has negative impacts. In Indonesia, based on data from the Cyber Crime Directorate (Dittipidsiber) website, the crime rate related to the ITE Law (Information and Electronic Transactions) is increasing day by day. This encourages digital forensic investigators to be able to develop a concept or method that can be adjusted to digital cases, for example cases of digital data manipulation such as photos or documents. Metadata is an information structure that describes, explains, places in a place or makes it easier to find something, use or manage and sources of information. Metadata can also be interpreted as data about data or information about information. One method or approach that can be done in cases of digital files (photos, videos or documents) can be done using forensic metadata analysis. This is because metadata stores information related to a file. By developing a library from java (metadata-extractor) based on open source and developed in the Netbeans 8.0 application, it will make it easier for an investigator or forensic investigator to conduct a forensic metadata approach, which is expected from the results can be used as valid evidence in the digital forensic investigation process. Keywords – Metadata, Digital Forensics, Cyber Crime, Metadata Analysis, Digital Evidence

The Impact of the Relationship between Company Characteristics and the Audit Committee on the Quality of Financial Reports (A Field Study on a Sample of Iraqi Private Companies)

The current study primarily aimed to determine the impact of the relationship between company characteristics and the audit committee on the quality of financial reports in its various dimensions (formal, temporal, and informational content). The study also sought to identify the influence of company characteristics on the quality of financial reports across these dimensions. The audit Committee role in improving financial report quality was the first area that the survey examined; this includes looking not just where audits. The study employed a descriptive and then analytical approach to achieve the goal of answering its research question. The study population included all senior management personnel at Iraqi private businesses, responsible auditors and members of audit commissions. From a variety of industries five companies were chosen, yielding a sample of 500 possible contacts (average 100 staff / company). A random sample of 219 participants was selected using the formula that Richard Geiger came up with. Questionnaires were sent to the participants, and all the responses accepted. After analyzing the data and testing the hypotheses using SPSS software, several key findings were revealed, the most significant being: • Company characteristics significantly affect the quality of financial reports. These characteristics provide an accurate and objective depiction of the company's financial position, aiding investors in making informed investment decisions. • Iraqi companies' unique characteristics notably affect the formal dimension of financial report information, as the company's reputation and business relationships are reflected in the reports' format, influencing investors' valuation of shares. • The timeliness of financial report information is vital in attracting investors. It reflects the company's growth and continuity in the market, significantly affecting its valuation. • Company characteristics influence the content of financial reports, including financial data and accompanying disclosures, thereby providing precise information for investors. • Audit committees play a pivotal role in ensuring the integrity and transparency of financial information, enhancing trust, and reducing errors in financial reports. They also contribute to improving the organizational structure of financial information, facilitating data comparability across different periods and entities. • The results indicate that audit committees improve the accuracy of the temporal information presented, boosting the reliability of information provided to the public. Additionally, audit committees assure investors that financial data are free from material errors, positively affecting investment decisions. • The research has shown that the relationship between company characteristics and audit committees enhances the verifiability of financial reports and also builds investor confidence. However, this research finds that if committee independence and effectiveness are not sufficiently so a very close relationship between enterprises (or groups of companies) and audit committees may have undesirable effects on quickness in presenting accountings. • The findings underscore that the unique characteristics of a company directly influence the content of its financial reports, highlighting the importance of having influential audit committees to ensure the credibility of the presented information.

Class-Imbalanced Machinery Fault Diagnosis Using Heterogeneous Data Fusion Support Tensor Machine

The monitoring signals of bearings from single-source sensor often contain limited information for characterizing various working condition, which may lead to instability and uncertainty of the class-imbalanced intelligent fault diagnosis. On the other hand, the vectorization of multi-source sensor signals may not only generate high-dimensional vectors, leading to increasing computational complexity and overfitting problems, but also lose the structural information and the coupling information. This paper proposes a new method for class-imbalanced fault diagnosis of bearing using support tensor machine (STM) driven by heterogeneous data fusion. The collected sound and vibration signals of bearings are successively decomposed into multiple frequency band components to extract various time-domain and frequency-domain statistical parameters. A third-order heterogeneous feature tensor is designed based on multi sensors, frequency band components and statistical parameters. STM based intelligent model is constructed to preserve the structural information of the third-order heterogeneous feature tensor for bearing fault diagnosis. A series of comparative experiments verify the advantages of the proposed method. Conflict of Interest Statement The authors declare no conflicts of interest.