Types Of Datasets Research Articles

Depositional-process controls on chemofacies in mixed-lithology submarine lobe deposits: a high-resolution core study from the Permian Wolfcamp XY Formation, Delaware Basin, Texas, U.S.A.

ABSTRACT Mixed siliciclastic–carbonate mudrocks have variable depositional processes and diagenetic pathways, creating mineralogical complexity and thus difficulty in characterizing reservoir quality using typical subsurface datasets (e.g., well logs) as well as conventional visual core-description techniques. Core-based X-ray fluorescence (XRF) data quantifies subtle elemental variations that can aid in interpreting fine-scale sedimentological packages and variations in reservoir-property distribution. XRF data has proven to be particularly useful for interpreting and defining the depositional processes of muddy, thin-bedded, mixed-lithology successions like the Wolfcamp and Bone Spring formations of the Delaware Basin, Texas, USA. These units consist of early to middle Permian siliciclastic and carbonate deep-marine deposits that form productive unconventional hydrocarbon reservoirs. However, the spatial and temporal variability in depositional processes and diagenetic evolution leads to difficulty in predicting reservoir presence and quality. Several studies have utilized core-based XRF data for this purpose, but not at a resolution sufficient to capture the true heterogeneity of these thin-bedded deposits. Using continuous, high-resolution (1 cm, 0.39 inch) X-ray fluorescence data from 66 m (218 feet) of core and associated geomechanical and well-log data from the Wolfcamp XY interval, this study demonstrates that chemofacies derived using unsupervised machine learning correlate with event-bed interpretations and reservoir-property distribution. Unsupervised k-means clustering and principal-component analysis on 17 XRF-derived elemental concentrations derive four chemofacies that characterize geochemical heterogeneity: 1) calcareous, 2) detrital, 3) oxic–suboxic argillaceous, and 4) anoxic argillaceous. The mineralogy and paragenesis of these chemofacies are validated using scanning-electron microscopy (SEM) and thin-section petrography. Vertical variations in XRF-based chemofacies accurately represent depositional facies changes and hybrid-event-bed boundaries, often matching cm-by-cm the visually-described lithofacies. We utilize this detailed dataset to construct a predictive chemofacies model linking variable sediment routing from carbonate and siliciclastic sources and various depositional processes to reservoir properties. This research also demonstrates that reservoir properties (e.g., total organic carbon, porosity, permeability, water saturation) and geomechanical response (brittleness and unconfined compressive strength) vary with chemofacies, with argillaceous facies generally being less brittle but having higher porosity. These results can be used for log-based reservoir prediction of the Wolfcamp and Bone Spring formations in the Permian Basin, as well as for other mixed siliciclastic–carbonate deep-water reservoirs around the world.

Author Correction: Typical and extreme weather datasets for studying the resilience of buildings to climate change and heatwaves

Author Correction: Typical and extreme weather datasets for studying the resilience of buildings to climate change and heatwaves

A Hybrid Intelligent Algorithm for Deterministic and Probabilistic Interval Predictions of a High-Speed Railway Wind Warning System

A wind warning system (WWS) is designed to ensure the operational safety of trains amid various wind environments. The system integrates the Internet, cloud computing, and virtual instrumentation technologies to automatically collect wind speed data and intelligently output alert information. To precisely predict the wind speed, a hybrid intelligent algorithm termed VMD–SSA–GRU comprising variational mode decomposition (VMD), sparrow search algorithm (SSA), and gated recurrent unit (GRU) is proposed and integrated into the WWS. Besides, the quantile regression (QR) is applied to wind speed uncertainty estimation, and the probability density functions of partial intervals are further evaluated by an improved kernel density estimation method. Two typical wind speed datasets (typhoon and normal wind conditions) measured by the WWS are used for validation of the applicability of the developed forecasted model. Through comparison to seven single and four combined models, the developed model presents the highest forecast accuracy in both definitive and uncertainty predictions even in typhoon wind conditions. The study demonstrates that the WWS integrated with the proposed hybrid intelligent algorithm can provide sensible warning for the safety of train operation.

Devising a method for detecting information threats in the Ukrainian cyber space based on machine learning

The object of this study is a disinformation detection process based on search algorithms for identifying fake news. The main task was to define a set of criteria and parameters for detecting the Ukrainian-language disinformation based on machine learning. A methodology has been considered for developing and filling a dataset of fakes for further training of the model and testing it for the purpose of identifying disinformation and propaganda, as well as determining the attributes of primary sources and routes of their distribution. This makes it possible to reasonably approach the definition of a model for forecasting the development of information threats in the cyberspace of Ukraine. In particular, the accuracy of automatic detection of the probability of disinformation in texts can be increased. For the English-language texts using balanced datasets for training when applying classical machine learning classifiers, the accuracy of identification and recognition of fakes is ³90 %, and for the Ukrainian-language texts – ³52 % and £90 %. That has made it possible to devise requirements for the structure and content of a typical dataset of fakes in the period after the full-scale invasion of Ukraine. The practical result of this work is the designed decision-making support system for monitoring, detecting, recognizing, and forecasting information threats in the cyberspace of Ukraine based on NLP and machine learning. The implementation of preliminary processing of the Ukrainian-language news, taking into account the linguistic features of the language in the text, increases the accuracy of fake identification by »1.72 times. Approaches to the construction of models for forecasting the development of information threats in cyberspace have been developed, which is an urgent task when fake news and information manipulation can affect public sentiment, politics, and the economy

Nanoparticles Ordering Classification Using Deep Convolutional Neural Networks

Intelligent neural networks are used efficiently for image classification and recognition in various scientific areas. One of the most important of these areas is nanoscience. Researchers are currently seeking to apply various deep learning neural networks models for fast and intelligent prediction and recognition of nanostructures based on scanning electron microscopy images. Models of Deep Convolutional Neural Networks (DCNN) have reached a high accuracy rate in nanoparticles classification and recognition. In fact, the improvement of the classification accuracy strongly relies on the perfect fine-tuning of image data and model parameters and that is what this research has worked for. The aim of this paper is to present a model, specifically the VGG16 convolutional neural network model, for high accurate nanoparticles ordering classification. The model has been used to classify the nanoparticles ordering using a typical dataset of electron microscopy images. In this research, an experiment has been carried out to achieve better accuracy rate in comparison to previously recorded accuracy rates. Data augmentation, modification techniques, and model tuning parameters are applied to excess the ability of the model for classifying the input image to ordered or non-ordered nanoparticles. Compared to the related works, the presented model has outperformed the pervious by achieving an accuracy rate of 97%. In this work, it has been observed that training iterations and balanced training data significantly improve the model performance and enhance the accuracy rate.

Recent Advances of Foundation Language Models-based Continual Learning: A Survey

Recently, foundation language models (LMs) have marked significant achievements in the domains of natural language processing (NLP) and computer vision (CV). Unlike traditional neural network models, foundation LMs obtain a great ability for transfer learning by acquiring rich commonsense knowledge through pre-training on extensive unsupervised datasets with a vast number of parameters. Despite these capabilities, LMs still struggle with catastrophic forgetting, hindering their ability to learn continuously like humans. To address this, continual learning (CL) methodologies have been introduced, allowing LMs to adapt to new tasks while retaining learned knowledge. However, a systematic taxonomy of existing approaches and a comparison of their performance are still lacking. In this paper, we delve into a comprehensive review, summarization, and classification of the existing literature on CL-based approaches applied to foundation language models, such as pre-trained language models (PLMs), large language models (LLMs) and vision-language models (VLMs). We divide these studies into offline and online CL, which consist of traditional methods, parameter-efficient-based methods, instruction tuning-based methods and continual pre-training methods. Additionally, we outline the typical datasets and metrics employed in CL research and provide a detailed analysis of the challenges and future work for LMs-based continual learning.

Cross-supervised contrastive learning domain adaptation network for steel defect segmentation

Defect segmentation is a vital approach to ensure the quality of steel. Convolutional neural network (CNN) has significantly enhanced steel defect segmentation performance. However, in practical scenario, steel defects exhibit great difference in appearance and background due to complex working conditions. It is challenging to perform steel defect segmentation under polytopic conditions. In this article, a cross-supervised contrastive learning domain adaptation network (CSCLDAN) is proposed for steel defect segmentation. Firstly, to alleviate domain shift posed by variations in defect appearance, a cross-supervised learning network that leverages CNN and Transformer is proposed to extract comprehensive features by facilitating the interaction of local and global information from different domains simultaneously. Secondly, for the domain shift caused by contextual difference, self-aware consistency contrastive learning is proposed to maintain consistency of the same defect features under different contexts. Finally, bidirectional pixel feature alignment loss is proposed to achieve self-aware consistency, which enables features from positive samples closer and separates features from negative samples. The effectiveness of CSCLDAN is verified on four typical steel defect datasets and the testing results demonstrate that CSCLDAN outperforms other state-of-the-art domain adaptation models.

SiamMask RAM for Underwater Target Tracking in Sonar Images

Abstract Target tracking based on sonar images plays a crucial role in underwater coordinated operations and strikes. In this paper, aiming to address the problem of underwater noise interference and target boundary blurring in sonar image, we propose an enhanced SiamMask network, SiamMask-RAM, for underwater target tracking. By combining hybrid attention and cross correlation structures, a new similarity metric branch is proposed to enhance perception of underwater target boundaries. Furthermore, by combining the confidence scores of the positive samples with the IoU, a ranking loss optimization strategy is proposed to reduce the possibility of mismatch between the classification and regression. Exceptionally, we constructed an underwater typical target dataset based on sonar images. The evaluation results on the sonar dataset demonstrate that SiamMask-RAM achieves a tracking accuracy of 0.752, an expected average overlap of 0.362, which are 22.6% and 21.9% better than the SiamRPN tracking algorithm, respectively. These findings indicate that the proposed method exhibits high accuracy and robustness in underwater tracking scenarios involving sonar images.

3D‐TabNetHS: A hyperspectral image classification method based on improved interpretable 3D attentive TabNet

AbstractThe classification methods for hyperspectral images (HSI) based on decision trees and convolutional neural networks have shown increasing advantages, but these methods often require a large number of labelled samples for learning, which is difficult for HSI, and the interpretability of the network is not high. Therefore, this paper proposes classification methods based on improved attention interpretable table learning (TabNet) named 3D TabNet HSI (3D‐TabNetHS) and unsupervised 3D TabNet HSI (U3D‐TabNetHS). These methods use sequential attention to select appropriate HSI spatial‐spectral features and add a space spectral information extraction (SSE) module composed of a 3D convolutional neural network (3D‐CNN) and fully connected layers to the Attention Transformer module in the original TabNet network to extract spatial‐spectral soft features. At the same time, unsupervised learning can be used to retrain the 3D‐TabNetHS network, and the classification accuracy of the resulting U3D‐TabNetHS network can be further improved. Compared with other HSI classification methods based on decision trees, the HSI classification accuracy of 3D‐TabNetHS is higher. On three typical HSI datasets, the accuracy metric overall accuracy of 3D‐TabNetHS reached as high as 98.71%, 94.73%, and 97.23%, respectively. Simultaneously, the consistency evaluation metric Kappa also reached 98.56%, 93.98%, and 96.31% respectively. The experimental results indicate the feasibility and reliability of the proposed method in HSI classification.

Semi-supervised convolutional generative adversarial networks for dynamic fault classification with manifold regularization

Data-driven fault classification identifies and categorizes faults or anomalies in industrial systems, to ensure efficient and safe operations. This paper addresses the challenges of semi-supervised learning for dynamic fault classification in industrial processes. The key problems involve utilizing three unique characteristics of typical industrial process dataset: 1) the presence of unlabeled samples due to high labeling costs and required expert knowledge, 2) the local manifold structure arising from strong variable coupling, and 3) time correlation inherent in time-series sensing data of dynamic processes. To tackle these challenges, a semi-supervised fault classification model, SSDCGAN, is proposed, based on deep convolutional Generative Adversarial Networks. The model captures dynamic temporal information through a moving window technique and leverages manifold regularization to maintain classifier consistency along the manifold’s tangent direction. Evaluations on the Tennessee Eastman (TE) benchmark demonstrate that SSDCGAN enhances fault classification accuracy, outperforming current methods.

Exploring feature sparsity for out-of-distribution detection

Out-of-distribution (OOD) detection is a crucial problem in practice, especially, for the safe deployment of machine learning models in industrial settings. Previous work has used free energy as a score function and proposed a fine-tuning method that utilized OOD data in the training phase of the classification model, which achieves a higher performance on the OOD detection task compared with traditional methods. One key drawback, however, is that the loss function parameters are highly dependent on involved datasets, which means it cannot be dynamically adapted and implemented in others settings; in other words, the general ability of the energy score is considerably limited. In this work, our point of departure is to enlarge distinguishability between in-distribution features and OOD data. Consequently, we present a simple yet effective sparsity-regularized (SR) tuning framework for this purpose. Our framework has two types of workflows depending on if external OOD data is available, the complexity of the original training loss is sharply reduced by adopting this modification, meanwhile, the adapted ability and detection performance are enhanced. Also, we contribute a mini dataset as a light and efficient alternative of the previous large-scale one. In the experiments, we verify the effectiveness of our framework in a wide range of typical datasets along with common network architectures.

Time domain correlation entropy image conversion: A new method for fault diagnosis of vehicle-mounted cable terminals

The identification of partial discharge (PD) in cable terminals is crucial for the safe operation of trains. However, the complexity of the operational environment and the similarity of PD signals make defect identification challenging. Consequently, this paper proposes a Time-domain Local Correlation Entropy Image (T-LCEI) transformation method, which constructs an entropy matrix to convert raw PD signals into images. These images embed feature and bandwidth information from the original PD data, significantly enhancing the ability to differentiate between similar PD signals. Furthermore, the method combines a Dual Attention Convolutional Neural Network (DA_CNN) for the effective classification of correlation entropy images. Experimental results demonstrate that this approach achieves an average classification accuracy of 99.69% across four typical PD defect datasets, with a testing accuracy of 97.75% in practical scenarios. Compared to existing PD detection methods, T-LCEI offers significant improvements in effectiveness and discriminability. The integration of DA_CNN further enhances recognition accuracy. The study demonstrates that the proposed method excels in PD defect identification, providing reliable technical support for on-site fault detection and maintenance, thereby significantly improving the operational safety of cable terminals.

Multi-population coevolutionary algorithm for a green multi-objective flexible job shop scheduling problem with automated guided vehicles and variable processing speed constraints

This study focuses on addressing a multi-objective Flexible Job Shop Scheduling Problem with Automated Guided Vehicles (FJSP-AGVs) and variable processing speed constraints. First, a position-based mixed integer linear programming model (MILP) is proposed to optimize simultaneously the maximum completion time and the total energy consumption. Then, we decompose FJSP-AGVs into four interrelated subproblems and design a Multi-Population Coevolutionary Algorithm (MCEA) to solve them. In MCEA, (1) The effective encoding and decoding methods are used to accurately reflect the characteristics of the problem, and generate feasible scheduling solutions. (2) A multi-rule-based heuristic is proposed to enrich the diversity of four populations. (3) A disjunctive graph is constructed to depict and obtain the critical path(s). On this basis, (4) two cooperative evolution strategies based on critical paths are proposed to facilitate collaborative evolution between different populations and improve the global search capability of the algorithm. Furthermore, (5) a consumption reduction strategy is proposed by reducing the processing speed of operations on non-critical paths while ensuring that it does not affect the makespan. Finally, we validate the effectiveness of MCEA by GD, and IGD, and set coverage metrics on the four typical benchmark datasets. Based on the average GD (IGD) metric across 65 instances, MCEA shows reductions of 77.63% (93.60%), 95.30% (97.27%), and 96.17%(97.89%) relative to EHA, EMOEA, and mop-BRKGA, respectively. The set coverage metric, MCEA outperforms EHA, EMOEA, and mop-BRKGA in 59, 64, and 64 instances, respectively. These results clearly indicate that MCEA can solve the FJSP-AGVs with variable processing speed constraints.

Improved Bald Eagle Search Optimization Algorithm for Feature Selection in Classification

Feature selection serves as an effective way for decreasing the quantity of features within a dataset, which helps enhance the performance of classification in machine learning (ML). In this paper, we formulate a joint feature selection problem to reduce the number of selected features while improving the classification accuracy. We propose an improved bald eagle search (IBES) algorithm to solve the optimization problem. Specifically, the BES algorithm is enhanced by introducing the lévy flight mechanism in the selection phase to improve the global search capability of the algorithm. In addition, we adopt an adaptive weighting factor to balance the global and local search capabilities. Finally, a novel mutation mechanism incorporating Gaussian and differential mutation is proposed, which contributes to maintain the diversity of the population. Comparative experiments are conducted with six benchmark algorithms on eighteen typical datasets. The results analysis indicates that the proposed IBES algorithm can achieve higher classification accuracy with a small number of features.

HFS-CSR: A hierarchical feature selection method based on correlation and structural redundancy

Feature selection (FS) is a crucial step in data engineering, as it not only saves time and resources for model training but also improves the prediction performance. In many realistic data sets, the feature space is formed by binary features which are related via generalization-specialization relationships (a.k.a., hierarchical feature space). These interdependencies among features, which can be described via a directed acyclic graph, present new challenges for classical FS methods. In this paper, a Hierarchical Feature Selection method based on Correlation and Structural Redundancy (HFS-CSR) is proposed to address it. In HFS-CSR, the correlation between features and the class label is measured based on Fisher Score whereas the structural redundancy among features are assessed quantitatively via hierarchical redundancy and depth penalties. The importance of features are determined by combining them adaptively for feature filtering. Experiments are conducted based on two kinds of typical data sets with hierarchical feature space: aging gene data sets and NLP data sets. The results demonstrate that HFS-CSR can achieve better performance compared with seven existing FS methods.

Cross-Modal Vertical Federated Learning for MRI Reconstruction.

Federated learning enables multiple hospitals to cooperatively learn a shared model without privacy disclosure. Existing methods often take a common assumption that the data from different hospitals have the same modalities. However, such a setting is difficult to fully satisfy in practical applications, since the imaging guidelines may be different between hospitals, which makes the number of individuals with the same set of modalities limited. To this end, we formulate this practical-yet-challenging cross-modal vertical federated learning task, in which data from multiple hospitals have different modalities with a small amount of multi-modality data collected from the same individuals. To tackle such a situation, we develop a novel framework, namely Federated Consistent Regularization constrained Feature Disentanglement (Fed-CRFD), for boosting MRI reconstruction by effectively exploring the overlapping samples (i.e., same patients with different modalities at different hospitals) and solving the domain shift problem caused by different modalities. Particularly, our Fed-CRFD involves an intra-client feature disentangle scheme to decouple data into modality-invariant and modality-specific features, where the modality-invariant features are leveraged to mitigate the domain shift problem. In addition, a cross-client latent representation consistency constraint is proposed specifically for the overlapping samples to further align the modality-invariant features extracted from different modalities. Hence, our method can fully exploit the multi-source data from hospitals while alleviating the domain shift problem. Extensive experiments on two typical MRI datasets demonstrate that our network clearly outperforms state-of-the-art MRI reconstruction methods.

Fall detection method based on spatio-temporal coordinate attention for high-resolution networks

Fall behavior is closely related to the high mortality rate of the elderly, so fall detection has become an important and urgent research area in human behavior recognition. However, the existing fall detection methods, suffer from the loss of detailed action information during feature extraction due to the downsampling operation, resulting in subpar performance when detecting falls with similar behaviors such as lying and sitting. To solve the challenges, this study proposes a high-resolution spatio-temporal feature extraction method based on a spatio-temporal coordinate attention mechanism. The method employs 3D convolutions to extract spatio-temporal features and utilizes gradual down-sampling to generate a multi-resolution sub-network, thus realizing multi-scale fusion and perception enhancement of details. In particular, this study designs a pseudo-3D basic block, which simulates the ability of 3D convolution, to ensure the running speed of the network while controlling the number of parameters. Further, a spatio-temporal coordinate attention mechanism is designed to accurately extract the spatio-temporal positional changes of key skeletal points and the interrelationships among them. Long-term dependencies in horizontal, vertical, temporal directions are captured through three one-dimensional global pooling operations. Then the long-range relationships and channel correlations among features are captured by cascading and slicing operations. Finally, the key information is effectively highlighted by performing dot-multiplication operations between the feature maps from the horizontal, vertical and temporal directions and the input feature maps. Experimental results on three typical public datasets show that the proposed method can better extract motion features and improve the accuracy of fall detection.

Class activation map-based slicing-concatenation and contrastive learning: A novel strategy for record-level atrial fibrillation detection

BackgroundDeep learning-based models for atrial fibrillation (AF) detection require extensive training data, which often necessitates labor-intensive professional annotation. While data augmentation techniques have been employed to mitigate the scarcity of annotated electrocardiogram (ECG) data, specific augmentation methods tailored for recording-level ECG annotations are lacking. This gap hampers the development of robust deep learning models for AF detection. MethodsWe propose a novel strategy, a combination of Class Activation Map-based Slicing-Concatenation (CAM-SC) data augmentation and contrastive learning, to address the current challenges. Initially, a baseline model incorporating a global average pooling layer is trained for classification and to generate class activation maps (CAMs), which highlight indicative ECG segments. After that, in each recording, indicative and non-indicative segments are sliced. These segments are subsequently concatenated randomly based on starting and ending Q points of QRS complexes, with indicative segments preserved to maintain label correctness. Finally, the augmented dataset undergoes contrastive learning to learn general representations, thereby enhancing AF detection performance. ResultsUsing ResNet-101 as the baseline model, training with the augmented data yielded the highest F1-score of 0.861 on the Computing in Cardiology (CinC) Challenge 2017 dataset, a typical AF dataset with recording-level annotations. The metrics outperform most previous studies. ConclusionsThis study introduces an innovative data augmentation method specifically designed for recording-level ECG annotations, significantly enhancing AF detection using deep learning models. This approach has substantial implications for future AF detection research.

A Review of Abnormal Crowd Behavior Recognition Technology Based on Computer Vision

Abnormal crowd behavior recognition is one of the research hotspots in computer vision. Its goal is to use computer vision technology and abnormal behavior detection models to accurately perceive, predict, and intervene in potential abnormal behaviors of the crowd and monitor the status of the crowd system in public places in real time, to effectively prevent and deal with public security risks and ensure public life safety and social order. To this end, focusing on the abnormal crowd behavior recognition technology in the computer vision system, a systematic review study of its theory and cutting-edge technology is conducted. First, the crowd level and abnormal behaviors in public places are defined, and the challenges faced by abnormal crowd behavior recognition are expounded. Then, from the dimensions based on traditional methods and based on deep learning, the mainstream technologies of abnormal behavior recognition are discussed, and the design ideas, advantages, and limitations of various methods are analyzed. Next, the mainstream software tools are introduced to provide a comprehensive reference for the technical framework. Secondly, typical abnormal behavior datasets at home and abroad are sorted out, and the characteristics of these datasets are compared in detail from multiple perspectives such as scale, characteristics, and uses, and the performance indicators of different algorithms on the datasets are compared and analyzed. Finally, the full text is summarized and the future development direction of abnormal crowd behavior recognition technology is prospected.

Our Model Achieves Excellent Performance on MovieLens: What Does It Mean?

A typical benchmark dataset for recommender system (RecSys) evaluation consists of user-item interactions generated on a platform within a time period. The interaction generation mechanism partially explains why a user interacts with (e.g., like, purchase, rate) an item, and the context of when a particular interaction happened. In this study, we conduct a meticulous analysis of the MovieLens dataset and explain the potential impact of using the dataset for evaluating recommendation algorithms. We make a few main findings from our analysis. First, there are significant differences in user interactions at the different stages when a user interacts with the MovieLens platform. The early interactions largely define the user portrait which affect the subsequent interactions. Second, user interactions are highly affected by the candidate movies that are recommended by the platform's internal recommendation algorithm(s). Third, changing the order of user interactions makes it more difficult for sequential algorithms to capture the progressive interaction process. We further discuss the discrepancy between the interaction generation mechanism that is employed by the MovieLens system and that of typical real-world recommendation scenarios. That is, the MovieLens dataset records \(\langle user-MovieLens\rangle\) interactions, but not \(\langle user-movie\rangle\) interactions. All research articles using the MovieLens dataset model the \(\langle user-MovieLens\rangle\) rather than the \(\langle user-movie\rangle\) interactions, making their results less generalizable to many practical recommendation scenarios in real-world settings. In summary, the MovieLens platform demonstrates an efficient and effective way of collecting user preferences to address cold-starts. However, models that achieve excellent recommendation accuracy on the MovieLens dataset may not demonstrate superior performance in practice , for at least two kinds of differences: (1) the differences in the contexts of user-item interaction generation, and (2) the differences in user knowledge about the item collections. While results on MovieLens can be useful as a reference, they should not be solely relied upon as the primary justification for the effectiveness of a recommendation system model.