Structural Classification Research Articles

Structural damage identification and classification from time-frequency impact echo data using deep learning

ABSTRACT Deep learning advancements have enabled non-destructive crack detection, yet it remains inadequate to quantitatively identify and evaluate the structural damage states leveraging deep learning with impact echo (IE). Therefore, the present study proposes a method to achieve this target on concrete structures, utilising Convolutional Neural Networks (CNNs) and IE. During the experiments, a reinforced concrete beam was loaded to simulate various damage levels. The IE test was conducted on the pure bending zone of the beam, and the obtained data were transformed into two-dimensional (2D) time-frequency data for a six-state damage dataset. Subsequently, several CNNs were used for training, testing, and analysing their performance. The findings revealed the networks’ proficiency in distinguishing various degrees of damage. Among them, GoogLeNet emerged as the most accurate classifier. Further analysis indicated that GoogLeNet, trained with datasets from at least two monitoring units, significantly outperformed those trained using datasets from a single unit, achieving a remarkable F1 score of no less than 0.778. Additionally, the study compared 1D and 2D GoogLeNet models trained on different data formats. The results showed that the model trained on 2D time-frequency data can achieve a higher accuracy of 0.984, surpassing the 1D model trained on time-series data.

Analisa Data Penjualan Pada Apotek Ritonga Farma Menggunakan Data Mining Apriori

A pharmacy is a place or business that is specifically dedicated to providing medicines and other health products to the public. This place is also known as a drugstore or drug store in some countries. Pharmacies provide medicines both prescribed by doctors and over-the-counter (over-the-counter), to help patients cope with health problems they are experiencing.Some pharmacies also offer additional services such as blood pressure checks, vaccinations, simple health checks, and health counseling to the public. In applying a priori methods to pharmacy, a deep understanding of data structure and proper product classification is needed to overcome this problem. By knowing the pattern of frequent purchases, pharmacies can place items that are often purchased together close together on shelves or strategic locations. This can increase the convenience of buyers and speed up the purchase process. A priori methods are techniques in data mining that are used to find hidden patterns or associations in large datasets. A priori methods look for relationships between items in a dataset that often appear together. The main principle of the a priori method is that if an item-set appears frequently together, then it is likely that the item-set will also appear frequently together in other transactions.

Exploring vision transformers and XGBoost as deep learning ensembles for transforming carcinoma recognition

Early detection of colorectal carcinoma (CRC), one of the most prevalent forms of cancer worldwide, significantly enhances the prognosis of patients. This research presents a new method for improving CRC detection using a deep learning ensemble with the Computer Aided Diagnosis (CADx). The method involves combining pre-trained convolutional neural network (CNN) models, such as ADaRDEV2I-22, DaRD-22, and ADaDR-22, using Vision Transformers (ViT) and XGBoost. The study addresses the challenges associated with imbalanced datasets and the necessity of sophisticated feature extraction in medical image analysis. Initially, the CKHK-22 dataset comprised 24 classes. However, we refined it to 14 classes, which led to an improvement in data balance and quality. This improvement enabled more precise feature extraction and improved classification results. We created two ensemble models: the first model used Vision Transformers to capture long-range spatial relationships in the images, while the second model combined CNNs with XGBoost to facilitate structured data classification. We implemented DCGAN-based augmentation to enhance the dataset’s diversity. The tests showed big improvements in performance, with the ADaDR-22 + Vision Transformer group getting the best results, with a testing accuracy of 93.4% and an AUC of 98.8%. In contrast, the ADaDR-22 + XGBoost model had an AUC of 97.8% and an accuracy of 92.2%. These findings highlight the efficacy of the proposed ensemble models in detecting CRC and highlight the importance of using well-balanced, high-quality datasets. The proposed method significantly enhances the clinical diagnostic accuracy and the capabilities of medical image analysis or early CRC detection.

Ensemble graph neural networks for structural classification of HIV inhibiting molecules

Ensemble graph neural networks for structural classification of HIV inhibiting molecules

Unsupervised learning modeling of the impact of Black Swan events on financial network reconfiguration: New insights from the COVID-19 outbreak and the Russia-Ukraine war

The emergence of Black Swan events, such as the COVID-19 pandemic and the Russian-Ukrainian war, has presented new challenges for stock markets and necessitates a redesign of portfolio management strategies. This work investigates the impact of unexpected and rare events, like the COVID-19 pandemic and the military conflict between Russia and Ukraine, on the structure of the cross-correlation network of stock returns. This research area has not yet been explored in depth. Previous research has primarily focused on the effects on stock returns, volatility, or changes in the topological properties of complex financial networks. In contrast, this study introduces a novel two-phase approach, integrating complex networks and self-organizing maps (the Kohonen network), to analyze changes in network structure and stock classification. A total of 483 US stocks from the S&P 500 index were analyzed. The findings reveal a notable contrast between the networks constructed before and after the onset of the COVID-19 pandemic (895 quotations for each network), along with significant disparities in stock classification according to centrality measures. Moreover, in-depth analysis has revealed significant distinctions between the SOM classification and the inherent partitioning based on economic sectors for both periods. Our research highlights the process of asset reconfiguration within the cross-correlation network, triggered by uncommon and unexpected events that disrupt financial market quotations. These findings are crucial for investors, portfolio managers, policymakers, and researchers.

Tectonically deformed coal: Focus on microstructures & implications for basin evolution

Traditionally, the Gondwana basins of India were believed to have formed primarily through extensional tectonics that created rift basins. However, recent models propose that strike-slip movements along deep-seated crustal fractures also contributed to their development. This article investigates this theory through a comprehensive structural analysis of the Bokaro Basin, located in the Damodar Valley of eastern India. This study deciphers tectonic history and evolutionary mechanisms of the Bokaro Basin by analyzing its coal samples, especially their structures—a relatively unexplored aspect in Indian coal research. We investigate maceral types, deformation manifestations, and fracture patterns at various scales. This led us to refine the structural classification of (Indian) coals. Deformation imprints were identified using remote sensing for regional features, geological fieldwork for megascopic observations, and petrographic analysis for micro-scale studies. The study utilized ArcGIS for lithology and fault mapping in conjunction with fieldwork in the highly deformed southwestern Bokaro Basin. Petrographic analysis revealed coal macerals and deformation signs. This research confirms that the Bokaro Basin's tectonic evolution involved uneven fault distribution, with significant rift and pull-apart mechanism. Coal seams show diverse deformation patterns, from ductile in fusinite to brittle in collotelinite, indicating multiple deformation stages under varying stress regimes. The East Bokaro Basin, with deeper seams, shows stronger deformation than the West. Shear stress, alongside extension, played a role in shaping the basin, with sigmoidal tension gashes from the coals of Barakar Formation, confirming active shearing through or afterwards coalification during or/and after the Early Permian time. The structural classification aids in coal bed methane (CBM) reservoir characterization, highlighting promising potential in cataclastic/tectonically deformed coal reservoirs. Tectonic deformation has affected the coals from the Bokaro Basin, resulting in structures that range from blocky to cataclastic, with some instances of granulation.

ECOD: integrating classifications of protein domains from experimental and predicted structures.

The evolutionary classification of protein domains (ECOD) classifies protein domains using a combination of sequence and structural data (http://prodata.swmed.edu/ecod). Here we present the culmination of our previous efforts at classifying domains from predicted structures, principally from the AlphaFold Database (AFDB), by integrating these domains with our existing classification of PDB structures. This combined classification includes both domains from our previous, purely experimental, classification of domains as well as domains from our provisional classification of 48 proteomes in AFDB predicted from model organisms and organisms of concern to global health. ECOD classifies over 1.8 M domains from over 1000 000 proteins collectively deposited in the PDB and AFDB. Additionally, we have changed the F-group classification reference used for ECOD, deprecating our original ECODf library and instead relying on direct collaboration with the Pfam sequence family database to inform our classification. Pfam provides similar coverage of ECOD with family classification while being more accurate and less redundant. By eliminating duplication of effort, we can improve both classifications. Finally, we discuss the initial deployment of DrugDomain, a database of domain-ligand interactions, on ECOD and discuss future plans.

CATH v4.4: major expansion of CATH by experimental and predicted structural data.

CATH (https://www.cathdb.info) is a structural classification database that assigns domains to the structures in the Protein Data Bank (PDB) and AlphaFold Protein Structure Database (AFDB) and adds layers of biological information, including homology and functional annotation. This article covers developments in the CATH classification since 2021. We report the significant expansion of structural information (180-fold) for CATH superfamilies through classification of PDB domains and predicted domain structures from the Encyclopedia of Domains (TED) resource. TED provides information on predicted domains in AFDB. CATH v4.4 represents an expansion of ∼64 844 experimentally determined domain structures from PDB. We also present a mapping of ∼90 million predicted domains from TED to CATH superfamilies. New PDB and TED data increases the number of superfamilies from 5841 to 6573, folds from 1349 to 2078 and architectures from 41 to 77. TED data comprises predicted structures, so these new folds and architectures remain hypothetical until experimentally confirmed. CATH also classifies domains into functional families (FunFams) within a superfamily. We have updated sequences in FunFams by scanning FunFam-HMMs against UniProt release 2024_02, giving a 276% increase in FunFams coverage. The mapping of TED structural domains has resulted in a 4-fold increase in FunFams with structural information.

Joint Classification of Hyperspectral and LiDAR Data via Multiprobability Decision Fusion Method

With the development of sensor technology, the sources of remotely sensed image data for the same region are becoming increasingly diverse. Unlike single-source remote sensing image data, multisource remote sensing image data can provide complementary information for the same feature, promoting its recognition. The effective utilization of remote sensing image data from various sources can enhance the extraction of image features and improve the accuracy of feature recognition. Hyperspectral remote sensing (HSI) data and light detection and ranging (LiDAR) data can provide complementary information from different perspectives and are frequently combined in feature identification tasks. However, the process of joint use suffers from data redundancy, low classification accuracy and high time complexity. To address the aforementioned issues and improve feature recognition in classification tasks, this paper introduces a multiprobability decision fusion (PRDRMF) method for the combined classification of HSI and LiDAR data. First, the original HSI data and LiDAR data are downscaled via the principal component–relative total variation (PRTV) method to remove redundant information. In the multifeature extraction module, the local texture features and spatial features of the image are extracted to consider the local texture and spatial structure of the image data. This is achieved by utilizing the local binary pattern (LBP) and extended multiattribute profile (EMAP) for the two types of data after dimensionality reduction. The four extracted features are subsequently input into the corresponding kernel–extreme learning machine (KELM), which has a simple structure and good classification performance, to obtain four classification probability matrices (CPMs). Finally, the four CPMs are fused via a multiprobability decision fusion method to obtain the optimal classification results. Comparison experiments on four classical HSI and LiDAR datasets demonstrate that the method proposed in this paper achieves high classification performance while reducing the overall time complexity of the method.

An annual land cover dataset for the Baltic Sea Region with crop types and peat bogs at 30 m from 2000 to 2022

We present detailed annual land cover maps for the Baltic Sea region, spanning more than two decades (2000–2022). The maps provide information on eighteen land cover (LC) classes, including eight general LC types, eight major crop types and grassland, and two peat bog-related classes. Our maps represent the first homogenized annual dataset for the region and address gaps in current land use and land cover products, such as a lack of detail on crop sequences and peat bog exploitation. To create the maps, we used annual multi-temporal remote sensing data combined with a data encoding structure and deep learning classification. We obtained the training data from publicly available open datasets. The maps were validated using independent field survey data from the Land Use/Cover Area Frame Survey (LUCAS) and expert annotations from high-resolution imagery. The quantitative and qualitative results of the maps provide a reliable data source for monitoring agricultural transformations, peat bog exploitation, and restoration activities in the Baltic Sea region and its surrounding countries.

Interhypothalamic adhesions: prevalence, structure, and location-based classification map in pediatric patients undergoing MRI.

Interhypothalamic adhesions (IHAs) have been reported only in the pediatric population, with unknown prevalence and histological composition. We aim to demonstrate their prevalence, assess their persistence through longitudinal imaging, classify IHAs by anatomical distribution, explore their structure, and report associated pathologies. A retrospective review was conducted on consecutive pediatric brain MRI studies obtained between January 2012, and December 2013. The presence of an IHA was only confirmed when observed on at least two planes. For each IHA, cross-sectional area was calculated, and signal intensities were measured at the center on sagittal T2WIs. Signal intensities were also measured in both cerebral white matter and gray matter for normalization and comparison. Patient demographics and clinical information were collected from electronic charts. Out of 1550 patients (0-17.9 years), 33 (19 males, 14 females) had an IHA, resulting in a 2.13% prevalence. Follow-up images were available for 19 IHA-positive patients, and IHAs were again seen in 92% of the follow-up scans (71/77). Normalized IHA signal highly correlated with normalized gray matter signal (r = 0.83, P < 0.001), but not with normalized white matter signal (r = -0.16, p = 0.494). Common co-occurring pathologies included hydrocephalus (n = 9), prematurity (n = 8), and corpus callosum abnormalities (n = 7). All type 3 IHAs (3/3) were accompanied by pituitary pathologies. IHAs have a prevalence of 2.13% in our cohort, and the majority persist in longitudinal studies. They showed gray matter signal intensity and Type 3 IHAs exclusively accompanied pituitary abnormalities.

An Exploration of Key Phrase Frames in Business Management Discussions for EAP Teaching

The interest in the phraseological nature of language generated many studies on phraseology in different genres. The focus of the current study is on key phrase frames in business management research article discussions. Key phrase frames are recurrent phraseological expressions with a variable slot and are specific to the studied context, for example genre or discipline. The analysis of the study is based on a corpus of research article discussions, extracted from six leading journals in business management, published between 2017 and 2021. Four-word phrase frames in the corpus were searched using the KfNgram program and refined based on the exclusion criteria to generate a list of phrase frames that was compared with the Corpus of Contemporary American English.The comparison yielded a total of 63 four-word key phrase frames in business management that were analyzed following Biber and Gray's (2013) structural classification and Simpson-Vlach and Ellis's (2010) functional taxonomy. In terms of structure, most key phrase frames were found to be non-verb content word frames while functionally, most serve referential discourse functions. The results suggest that phraseological variation in business management is associated with both lexical and grammatical constructions. These findings are crucial as they have implications on discipline-specific academic writing. most key phrase frames were found to be non-verb content word frames while functionally, most serve referential discourse functions.The results suggest that phraseological variation in business management is associated with both lexical and grammatical constructions. These findings are crucial as they have implications on discipline-specific academic writing. most key phrase frames were found to be non-verb content word frames while functionally, most serve referential discourse functions. The results suggest that phraseological variation in business management is associated with both lexical and grammatical constructions. These findings are crucial as they have implications on discipline-specific academic writing.

Definition of reproductive structures in Eucalyptus for phenological data collection.

In an era where global climate change is shifting plant phenology, global meta-analyses of multiple species are required more than ever. Common language or references for enhanced data compatibility are key for such analyses. Although the Plant Phenology Ontology (PPO) addresses this challenge, it does not capture several relevant reproductive structures that are critical in species with long reproductive cycles, like many Eucalyptus species. We reviewed the terminology and concepts that describe reproductive structures in eucalypts and compared them with the existing classes of the PPO to explore the PPO's potential for harmonizing disparate eucalypt datasets. We identified incongruencies within and between eucalypt terminology and the PPO. We tested the sensitivity of the PPO for capturing key eucalypt phenological structures and found it sensitive to classification of certain structures. To address these limitations, we developed the Eucalyptus Phenology Ontology (EPO), a new ontology that builds on the PPO and captures key reproductive structures using a more refined classification. The EPO integrates the relationships between reproductive structures, phenological stages, and phenological traits. The vocabulary is species-neutral so it can be applied to other taxa but specifies the synonyms and descriptions required to capture the complexity of eucalypt phenology.

The Pfam protein families database: embracing AI/ML.

The Pfam protein families database is a comprehensive collection of protein domains and families used for genome annotation and protein structure and function analysis (https://www.ebi.ac.uk/interpro/). This update describes major developments in Pfam since 2020, including decommissioning the Pfam website and integration with InterPro, harmonization with the ECOD structural classification, and expanded curation of metagenomic, microprotein and repeat-containing families. We highlight how AlphaFold structure predictions are being leveraged to refine domain boundaries and identify new domains. New families discovered through large-scale sequence similarity analysis of AlphaFold models are described. We also detail the development of Pfam-N, which uses deep learning to expand family coverage, achieving an 8.8% increase in UniProtKB coverage compared to standard Pfam. We discuss plans for more frequent Pfam releases integrated with InterPro and the potential for artificial intelligence to further assist curation. Despite recent advances, many protein families remain to be classified, and Pfam continues working toward comprehensive coverage of the protein universe.

Fracture mode classification of UHPC based on deep learning and wavelet time–frequency spectrum derived from acoustic emission waveform

Fracture mode categorization is essential in identifying the failure stage, evaluating the damage characteristics, and providing references for structural maintenance. Acoustic emission (AE), as a passive nondestructive technique, presents great potential in damage characterization of cementitious materials and structures. In this study, a convolutional neural network (CNN) and gated recurrent unit (GRU) integrated model for tensile, shear, and mixed cracking mode classification of ultra-high-performance concrete (UHPC) was proposed and verified based on AE waveform features. First, four-point bending and direct shear tests for UHPC specimens were carried out, respectively, to generate AE waveform signals corresponding to different cracking modes (tensile, shear, and mixed ones). Subsequently, the collected one-dimensional AE signals were converted into two-dimensional time–frequency spectrum by continuous wavelet transform. And an AE dataset was constructed after manual labeling based on the corresponding relationships between cracking modes and wavelet time–frequency spectra. Then, an automatic CNN-GRU automatic classification model was established using wavelet time–frequency spectrum images as input variables, where the CNN extracts spatial image features, and the GRU screens time dependencies and realizes final classification. The proposed CNN-GRU model achieves an accuracy of 96.17% for cracking mode classification of UHPC, which is superior both in computational accuracy and efficiency to lightweight CNN and AlexNet. Classification outcomes for UHPC specimen under four-point bending and direct shear tests revealed the proportion of tensile cracks progressively decreased with the increase of damage, while shear cracks exhibited the opposite trend. The percentage of mixed mode cracks maintained relatively stable throughout the failure stages. The model proposed in this study presents favorable performance to reveal the underlying damage evolution mechanism during UHPC fracture, which can also provide references for the cracking mode classification of other cementitious materials and structures.

Latin American Natural Product Database (LANaPDB): An Update.

Natural product (NP) databases are crucial tools in computer-aided drug design (CADD). Over the past decade, there has been a worldwide effort to assemble information regarding natural products (NPs) isolated and characterized in certain geographical regions. In 2023, it was published LANaPDB, and to our knowledge, this is the first attempt to gather and standardize all the NP databases of Latin America. Herein, we present and analyze in detail the contents of an updated version of LANaPDB, which includes 619 newly added compounds from Colombia, Costa Rica, and Mexico. The present version of LANaPDB has a total of 13 578 compounds, coming from ten databases of seven Latin American countries. A chemoinformatic characterization of LANaPDB was carried out, which includes the structural classification of the compounds, calculation of six physicochemical properties of pharmaceutical interest, and visualization of the chemical space by employing and comparing two different fingerprints (MACCS keys (166-bit) and Morgan2 (2048-bit)). Furthermore, additional analyses were made, and valuable information not included in the first version of LANaPDB was added, which includes structural diversity, molecular complexity, synthetic feasibility, commercial availability, and reported and predicted biological activity. In addition, the LANaPDB compounds were cross-referenced to two of the largest public chemical compound databases annotated with biological activity: ChEMBL and PubChem.

3D dataset generation using virtual reality for forest biodiversity

ABSTRACT While forest biodiversity faces a concerning decline, modern technology presents promising avenues for mitigation. However, a critical gap persists in reconciling ecological knowledge with the technical expertise required to use state-of-the-art technologies in 3D data classification. Currently, one main issue is the scarcity of 3D datasets for biodiversity, particularly within the context of machine learning applications. Unlike the straightforward classification of human-made structures, forest environments are uniquely intricate and nuanced due to its inherently complex nature. This study addresses this challenge by introducing a fully automated pipeline for tree stem 3D point cloud segmentation, focussing on a biodiversity indicator: tree-related microhabitats (TreMs). Furthermore, our research advances the field by demonstrating that machine learning models trained with labels generated by our proposed virtual reality (VR) method, Labelling Flora, yield predictions statistically similar to the traditional desktop-based labelling methods. This implies that existing 3D datasets could be augmented using the more rapid approach of VR labelling. Additionally, the findings of this paper demonstrate the potential integration of VR and immersive technology into the 3D labelling workflow, facilitating a quicker and more intuitive labelling process. This could empower users, who are non-familiar with 3D modelling, to contribute their expertise to the segmentation process.

Research Progress of Natural Products with the Activity of Anti-nonalcoholic Steatohepatitis.

Nonalcoholic steatohepatitis (NASH), a multi-target disease, is becoming a global epidemic. Although several anti-NASH drug candidates are being evaluated in late-stage clinical trials, none have been approved by the FDA to date. Given the global prevalence of the disease, the lack of effective drugs, and the very limited therapeutic efficacy of most of the existing synthetic drugs focusing on a single target, there is an urgent need to continue to develop new therapeutic agents. In contrast, many natural products, including pure compounds and crude extracts, possess hepatoprotective activities. Usually, these natural components are characterized by multi-targeting and low side effects. Therefore, natural products are important resources for the development of new anti- NASH drugs. In this paper, we focus on reviewing the anti-NASH potential, structure, and some of the side effects of natural products based on structural classification. We hope this mini-review will help researchers design and develop new anti-NASH drugs, especially based on the structure of natural products.

Monotonicity, asymptotics and level sets for principal eigenvalues of some elliptic operators with shear flow

We investigate the joint effects of diffusion and advection on principal eigenvalues of some elliptic operators with shear flow. Some monotonicity and asymptotic behaviors of principal eigenvalues, with respect to diffusion rate and flow amplitude, are established. These analyses lead to a classification of topological structures of level sets for principal eigenvalues, as a function of diffusion rate and flow amplitude. Our analytical results provide a unifying viewpoint to understand mixing enhancement and dispersal-induced growth, which are apparently two unrelated phenomena, one in fluid mechanics and the other in population dynamics. In our analysis, some limiting Hamilton-Jacobi equations, blowup argument and limiting generalized principal eigenvalue problems play critical roles.

Pore Engineering in Biomass-Derived Carbon Materials for Enhanced Energy, Catalysis, and Environmental Applications.

Biomass-derived carbon materials (BDCs) are highly regarded for their renewability, environmental friendliness, and broad potential for application. A significant advantage of these materials lies in the high degree of customization of their physical and chemical properties, especially in terms of pore structure. Pore engineering is a key strategy to enhance the performance of BDCs in critical areas, such as energy storage, catalysis, and environmental remediation. This review focuses on pore engineering, exploring the definition, classification, and adjustment techniques of pore structures, as well as how these factors affect the application performance of BDCs in energy, catalysis, and environmental remediation. Our aim is to provide a solid theoretical foundation and practical guidance for the pore engineering of BDCs to facilitate the rapid transition of these materials from the laboratory to industrial applications.