Features Of Dataset Research Articles

A Quantification Method for Micro-defect Size in Copper Foil Based on Motion-induced Eddy Current Signal Characteristics

Abstract Lithium batteries represent a pivotal technology in modern energy storage and have attracted significant attention. Copper foil is frequently employed as the current collector for the negative electrode; however, the presence of micro-defects can severely impair the electrochemical performance of lithium batteries. Therefore, developing a method that can accurately detect micro-defects during high-speed production processes and subsequently achieve precise quantification of defect sizes is of paramount importance. Motion-Induced Eddy Current (MIEC) techniques, characterized by relative motion, can detect defects in conductive materials during high-speed movement. However, no research has been conducted on defect size quantification based on motion-induced eddy current signals. Consequently, this study proposes a quantification method for micro-defect size in copper foil based on motion-induced eddy current signal characteristics. By combining wavelet soft-thresholding and morphological filtering, high-frequency noise and signal fluctuations caused by variations in lift-off distance are effectively removed from the measured signals. Additionally, eight different features are defined for the micro-defect motion-induced eddy current signals, and a feature dataset for various sizes of copper foil micro-defects is established under different motor speeds. The proposed IVY-CNN-BiLSTM-Attention model is then utilized for the quantification of micro-defect sizes, yielding excellent results.

Enhancing Software Effort Estimation with Pre-Trained Word Embeddings: A Small-Dataset Solution for Accurate Story Point Prediction

Traditional software effort estimation methods, such as term frequency–inverse document frequency (TF-IDF), are widely used due to their simplicity and interpretability. However, they struggle with limited datasets, fail to capture intricate semantics, and suffer from dimensionality, sparsity, and computational inefficiency. This study used pre-trained word embeddings, including FastText and GPT-2, to improve estimation accuracy in such cases. Seven pre-trained models were evaluated for their ability to effectively represent textual data, addressing the fundamental limitations of TF-IDF through contextualized embeddings. The results show that combining FastText embeddings with support vector machines (SVMs) consistently outperforms traditional approaches, reducing the mean absolute error (MAE) by 5–18% while achieving accuracy comparable to deep learning models like GPT-2. This approach demonstrated the adaptability of pre-trained embeddings for small datasets, balancing semantic richness with computational efficiency. The proposed method optimized project planning and resource allocation while enhancing software development through accurate story point prediction while safeguarding privacy and security through data anonymization. Future research will explore task-specific embeddings tailored to software engineering domains and investigate how dataset characteristics, such as cultural variations, influence model performance, ensuring the development of adaptable, robust, and secure machine learning models for diverse contexts.

Research on the relationship between feature extraction time and training samples of hyperspectral image based on spatial domain

Hyperspectral image (HSI) feature extraction is an important means to improve the classification of different ground features. According to the structural characteristics of hyperspectral data, the general feature extraction scheme can extract features from the point of view of spectral dimension, spatial and spatial spectrum. And the feature extraction time is also an index to measure the feature extraction method. Therefore, from the perspective of spatial dimension, this paper explores the relationship between HSI feature extraction time and training sample ratio. Three groups of HSIs sets were used for correlation test and analysis in the experiment. According to the characteristics of different data sets, the best selection scheme between spatial domain feature extraction method and training samples is given.

Manufacturing Feature Recognition with a Sparse Voxel-based Convolutional Neural Network

Abstract Automated manufacturing feature recognition is a crucial link between computer-aided design and manufacturing, facilitating process selection and other downstream tasks in computer-aided process planning. While various methods such as graph-based, rule-based, and neural networks have been proposed for automatic feature recognition, they suffer from poor scalability or computational inefficiency. Recently, voxel-based convolutional neural networks have shown promise in solving these challenges but incur a tradeoff between computational cost and feature resolution. This paper investigates a computationally efficient sparse voxel-based convolutional neural network for manufacturing feature recognition, specifically, an octree-based sparse voxel convolutional neural network. This model is trained on a large-scale manufacturing feature dataset, and its performance is compared to a voxel-based feature recognition model (FeatureNet). The results indicate that the octree-based model yields higher feature recognition accuracy (99.5% on the test dataset) with 44% lower GPU memory consumption than a voxel-based model of comparable resolution. In addition, increasing the resolution of the octree-based model enables recognition of finer manufacturing features. These results indicate that a sparse voxel-based convolutional neural network is a computationally efficient deep learning model for manufacturing feature recognition to enable process planning automation. Moreover, the sparse voxel-based neural network demonstrated comparable performance to a boundary representation-based feature recognition neural network, achieving similar accuracy in single feature recognition without having access to the exact 3D shape descriptors

One size does not fit all in evaluating model selection scores for image classification

Selecting pretrained models for image classification often involves computationally intensive finetuning. This study addresses a research gap in the standardized evaluation of transferability scores, which could simplify model selection by ranking pretrained models without exhaustive finetuning. The motivation is to reduce the computational burden of model selection through a consistent approach that guides practitioners in balancing accuracy and efficiency across tasks. This study evaluates 14 transferability scores on 11 benchmark datasets. It includes both Convolutional Neural Network (CNN) and Vision Transformer (ViT) models and ensures consistency in experimental conditions to counter the variability in previous research. Key findings reveal significant variability in score effectiveness based on dataset characteristics (e.g., fine-grained versus coarse-grained classes) and model architectures. ViT models generally show superior transferability, especially for fine-grained datasets. While no single score is best in all cases, some scores excel in specific contexts. In addition to predictive accuracy, the study also evaluates computational efficiency and identifies scores that are suitable for resource-constrained scenarios. This research provides insights for selecting appropriate transferability scores to optimize model selection strategies to facilitate efficient deployment in practice.

Comparative analysis of impact of classification algorithms on security and performance bug reports

Abstract Identification and classification of bugs, e.g., security and performance are a preemptive and fundamental practice which contributes to the development of secure and efficient software. Software Quality Assurance (SQA) needs to classify bugs into relevant categories, e.g., security and performance bugs since one type of bug may have a higher preference over another, thus facilitating software evolution and maintenance. In addition to classification, it would be ideal for the SQA manager to prioritize security and performance bugs based on the level of perseverance, severity, or impact to assign relevant developers whose expertise is aligned with the identification of such bugs, thus facilitating triaging. The aim of this research is to compare and analyze the prediction accuracy of machine learning algorithms, i.e., Artificial neural network (ANN), Support vector machine (SVM), Naïve Bayes (NB), Decision tree (DT), Logistic regression (LR), and K-nearest neighbor (KNN) to identify security and performance bugs from the bug repository. We first label the existing dataset from the Bugzilla repository with the help of a software security expert to train the algorithms. Our research type is explanatory, and our research method is controlled experimentation, in which the independent variable is prediction accuracy and the dependent variables are ANN, SVM, NB, DT, LR, and KNN. First, we applied preprocessing, Term Frequency-Inverse Document Frequency feature extraction methods, and then applied classification algorithms. The results were measured through accuracy, precision, recall, and F-measure and then the results were compared and validated through the ten-fold cross-validation technique. Comparative analysis reveals that two algorithms (SVM and LR) perform better in terms of precision (0.99) for performance bugs and three algorithms (SVM, ANN, and LR) perform better in terms of F1 score for security bugs as compared to other classification algorithms which are essentially due to the linear dataset and extensive number of features in the dataset.

Machine Learning-Based Summer Crops Mapping Using Sentinel-1 and Sentinel-2 Images

Accurate crop type mapping using satellite imagery is crucial for food security, yet accurately distinguishing between crops with similar spectral signatures is challenging. This study assessed the performance of Sentinel-2 (S2) time series (spectral bands and vegetation indices), Sentinel-1 (S1) time series (backscattering coefficients and polarimetric parameters), alongside phenological features derived from both S1 and S2 time series (harmonic coefficients and median features), for classifying sunflower, soybean, and maize. Random Forest (RF), Multi-Layer Perceptron (MLP), and XGBoost classifiers were applied across various dataset configurations and train-test splits over two study sites and years in France. Additionally, the InceptionTime classifier, specifically designed for time series data, was tested exclusively with time series datasets to compare its performance against the three general machine learning algorithms (RF, XGBoost, and MLP). The results showed that XGBoost outperformed RF and MLP in classifying the three crops. The optimal dataset for mapping all three crops combined S1 backscattering coefficients with S2 vegetation indices, with comparable results between phenological features and time series data (mean F1 scores of 89.9% for sunflower, 76.6% for soybean, and 91.1% for maize). However, when using individual satellite sensors, S1 phenological features and time series outperformed S2 for sunflower, while S2 was superior for soybean and maize. Both phenological features and time series data produced close mean F1 scores across spatial, temporal, and spatiotemporal transfer scenarios, though median features dataset was the best choice for spatiotemporal transfer. Polarimetric S1 data did not yield effective results. The InceptionTime classifier further improved classification accuracy over XGBoost for all crops, with the degree of improvement varying by crop and dataset (the highest mean F1 scores of 90.6% for sunflower, 86.0% for soybean, and 93.5% for maize).

Enhancing wave energy farm efficiency: Eigen-stacking ensemble framework

The significant wave height (SWH) plays a pivotal role across diverse domains, ranging from the assessment of wave energy potential to the optimization of marine operations and the advancement of ocean engineering techniques. Understanding the spatial and temporal distribution patterns of SWH data among buoy network stations holds utmost importance for various applications. This research introduces a novel methodology for accurate spatiotemporal SWH prediction across multiple stations within buoy monitoring networks. Leveraging eigen SWH time series enables the identification of dataset patterns, feature extraction, and dimensionality reduction. Examining eigen time series provides insights into SWH dynamics, allowing exclusive use for prediction across all stations. By incorporating this eigen SWH time series as the sole input into a stacking ensemble model, we introduce a highly efficient and accurate framework for SWH prediction. Notably, the methodology achieves success in forecasting hourly SWH values along the western United States coastline, utilizing a stacking ensemble technique for enhanced accuracy. Evaluation metrics confirm outstanding performance, with CE and R2 values surpassing 0.95 and 0.93, respectively. Further, a single Long Short-Term Memory (LSTM) model is incorporated as a benchmark to assess the effectiveness of the stacking ensemble model for significant wave height (SWH) prediction, with results demonstrating that the ensemble model outperforms the single LSTM in terms of accuracy and robustness. Consequently, the introduced stacking ensemble model promises autonomy in forecasting spatial-temporal SWH patterns, extending its applicability within ocean engineering and scientific research.

Label Propagation Algorithm for Face Clustering using Shared Nearest Neighbor Similarity

Facial image datasets are particularly vulnerable to challenges such as lighting variations and occlusion, which can complicate data classification. Semi-supervised learning, using a limited amount of labeled facial data, offers a solution by enhancing face classification accuracy while reducing manual labeling efforts. The Label Propagation Algorithm (LPA) is a commonly used semi-supervised algorithm that employs Radial Basis Function (RBF) to measure similarities between data nodes. However, RBF struggles to capture complex nonlinear relationships in facial data. To address this, an improved LPA is proposed that integrates Shared Nearest Neighbor (SNN) to enhance the correlation measurement between facial data and RBF. Three known datasets were considered: FERET, Yale, and ORL. The experiments showed that in the case of insufficient label samples, the accuracy reached 89.76%, 92.46%, and 81.48%, respectively. The proposed LPA enhances clustering robustness by introducing 128 dimensional facial features and more complex similarity measurement. The parameter of similarity measurement can be adjusted based on the characteristics of different datasets to achieve better clustering results. The improved LPA achieved better performance and face clustering effectiveness by enhancing robustness and adaptability.

Another pipeline in local Partial Least Squares Regression (LPLS) methods: Assessing the impact of wavelet transform integration

This study evaluates the performance of four Partial Least Squares Regression (PLS) methods, focusing on a new Local Partial Least Squares Regression (LPLS) variant integrating wavelet transformation, named WLPLS, for analyzing feed using Near-Infrared (NIR) spectroscopy. While traditional PLS methods are effective for many spectroscopic applications, their global modeling approach often reduces predictive accuracy in large, heterogeneous datasets. In contrast, LPLS adapts models to local data characteristics, which can enhance prediction but also increase computational demands (power and time). WLPLS seeks to mitigate these demands by incorporating wavelet transformation to reduce data dimensionality while effectively managing spectral variances. This research conducts a comparative analysis of WLPLS against traditional PLS, LPLS, and another LPLS pipeline reducing data dimensionality: the LPLS on global PLS scores (LPLS-S). The performances of these methods were evaluated using a large feed database containing 24,644 samples, analyzing five key constituents: ash, crude fibers, fat, moisture, and proteins. The results demonstrate that local approaches outperformed the global PLS method for this dataset and that the performances of the local methods were relatively similar to each other. The selection of the optimal method therefore depends on the specific requirements of the application, such as dataset characteristics and the required prediction speed. Future studies should broaden this comparative framework to additional datasets and contexts to ensure they are adapted for diverse applications.

Large language models as an academic resource for radiologists stepping into artificial intelligence research.

Radiologists increasingly use artificial intelligence (AI) to enhance diagnostic accuracy and optimize workflows. However, many lack the technical skills to effectively apply machine learning (ML) and deep learning (DL) algorithms, limiting the accessibility of these methods to radiology researchers who could otherwise benefit from them. Large language models (LLMs), such as GPT-4o, may serve as virtual advisors, offering tailored algorithm recommendations for specific research needs. This study evaluates GPT-4o's effectiveness as a recommender system to enhance radiologists' understanding and implementation of AI in research. GPT-4o was used to recommend ML and DL algorithms based on specific details provided by researchers, including dataset characteristics, modality types, data sizes, and research objectives. The model acted as a virtual advisor, guiding researchers in selecting the most appropriate models for their studies. The study systematically evaluated GPT-4o's recommendations for clarity, task alignment, model diversity, and baseline selection. Responses were graded to assess the model's ability to meet the needs of radiology researchers. GPT-4o effectively recommended appropriate ML and DL algorithms for various radiology tasks, including segmentation, classification, and regression in medical imaging. The model suggested a diverse range of established and innovative algorithms, such as U-Net, Random Forest, Attention U-Net, and EfficientNet, aligning well with accepted practices. GPT-4o shows promise as a valuable tool for radiologists and early career researchers by providing clear and relevant AI and ML algorithm recommendations. Its ability to bridge the knowledge gap in AI implementation could democratize access to advanced technologies, fostering innovation and improving radiology research quality. Further studies should explore integrating LLMs into routine workflows and their role in ongoing professional development.

Brain-phenotype predictions of language and executive function can survive across diverse real-world data: Dataset shifts in developmental populations

Predictive modeling potentially increases the reproducibility and generalizability of neuroimaging brain-phenotype associations. Yet, the evaluation of a model in another dataset is underutilized. Among studies that undertake external validation, there is a notable lack of attention to generalization across dataset-specific idiosyncrasies (i.e., dataset shifts). Research settings, by design, remove the between-site variations that real-world and, eventually, clinical applications demand. Here, we rigorously test the ability of a range of predictive models to generalize across three diverse, unharmonized developmental samples: the Philadelphia Neurodevelopmental Cohort (n=1291), the Healthy Brain Network (n=1110), and the Human Connectome Project in Development (n=428). These datasets have high inter-dataset heterogeneity, encompassing substantial variations in age distribution, sex, racial and ethnic minority representation, recruitment geography, clinical symptom burdens, fMRI tasks, sequences, and behavioral measures. Through advanced methodological approaches, we demonstrate that reproducible and generalizable brain-behavior associations can be realized across diverse dataset features. Results indicate the potential of functional connectome-based predictive models to be robust despite substantial inter-dataset variability. Notably, for the HCPD and HBN datasets, the best predictions were not from training and testing in the same dataset (i.e., cross-validation) but across datasets. This result suggests that training on diverse data may improve prediction in specific cases. Overall, this work provides a critical foundation for future work evaluating the generalizability of brain-phenotype associations in real-world scenarios and clinical settings.

Assessment of CO2 Emissions for Light-Duty Vehicles Using Dynamic Perturbation Additive Regression Trees

Effective predictive modeling is crucial for assessing and mitigating energy consumption and CO2 emissions in light-duty vehicles (LDVs) throughout the whole value chain of an organization. This study enhances the modeling of LDV CO2 emissions by developing novel approaches to analyzing vehicle feature datasets. New tree-based machine learning models are developed to increase the accuracy and interpretability in modeling the CO2 emissions in LDVs. In particular, this study develops a new algorithm called dynamic perturbation additive regression trees (DPART). This new algorithm integrates dynamic perturbation within an iterative boosting framework. DPART progressively adjusts prediction values and explores various tree structures to improve predictive performance with reduced computation time. The effectiveness of the new ensemble-tree-based models is compared to that of other models for the vehicle emission data. The results demonstrate the new models’ capability to significantly improve predicting accuracy and reliability compared to other models. The new models also enable identifying key vehicle features affecting emissions, and thus provide valuable insights into the complex relationships among vehicle features in the dataset.

Efficient data processing pipeline for event-based vision datasets: techniques and insights

Abstract Event-based vision datasets have emerged as a critical asset in advancing the capabilities of real-time perception systems, particularly in fields such as robotics, autonomous vehicles, and human-computer interaction. These datasets enable low-latency processing by capturing asynchronous, pixel-level changes in the scene, providing a distinct advantage over traditional frame-based systems. However, the diverse formats and characteristics of event-based datasets pose significant challenges for efficient processing and analysis, hindering their broader adoption and integration. In this paper, we present a versatile and comprehensive data processing pipeline designed to address these challenges by supporting multiple event-data formats, including newer formats such as EVT2 and EVT3. This pipeline not only converts data into widely supported formats like AEDAT and NPZ, but also ensures that the unique characteristics of event-based data-such as temporal precision and sparse event representation-are preserved throughout the conversion process. By applying this pipeline to several open-source datasets, we establish a standardized, efficient methodology for dataset manipulation that enhances compatibility and reproducibility in event-based vision research. Additionally, we introduce a novel high-resolution event-based action dataset, converted into various formats using our pipeline, which opens new avenues for exploring event-based techniques in action recognition. This dataset and our pipeline serve as valuable resources for the research community, enabling advancements in real-time vision applications and fostering greater collaboration and standardization across studies.

Utilization of Artificial Intelligence for the automated recognition of fine arts

Fine art recognition, traditionally dependent on human expertise, is undergoing a significant transformation with the integration of Artificial Intelligence (AI) and deep learning. This article introduces a novel AI-based approach for fine art recognition, utilizing Convolutional Neural Networks (CNNs) and advanced feature extraction techniques. Addressing the inherent challenges within this domain, we present a systematic methodology to enhance automated fine art recognition. By leveraging critical dataset characteristics such as objective type, genre, material, technique, and department, our method exhibits exceptional performance in classifying fine art pieces across diverse attributes. Our approach significantly improves accuracy and efficiency by integrating advanced feature extraction techniques with a customized CNN architecture. Experimental validation on a benchmark dataset highlights the efficacy of our method, indicating substantial contributions to the interdisciplinary field of fine art analysis.

Comparative Analysis of Machine Learning Algorithms for Cross-Site Scripting (XSS) Attack Detection

Cross-Site Scripting (XSS) attacks pose a significant cybersecurity threat by exploiting vulnerabilities in web applications to inject malicious scripts, enabling unauthorized access and execution of malicious code. Traditional XSS detection systems often struggle to identify increasingly complex XSS payloads. To address this issue, this research evaluated the efficacy of Machine Learning algorithms in detecting XSS threats within online web applications. The study conducts a comprehensive comparative analysis of XSS attack detection using four prominent Machine Learning algorithms, which consist of Extreme Gradient Boosting (XGBoost), Random Forest (RF), K-Nearest Neighbors (KNN), and Support Vector Machine (SVM). This research utilizes a comparative methodology to assess the selected Machine Learning algorithms by analyzing their performance metrics, including confusion matrix, 10-fold cross-validation, and assessment of training time to thoroughly evaluate the models. By exploring dataset characteristics and evaluating the performance metrics of each selected algorithm, the study determined the most robust Machine Learning solution for XSS detection. Results indicate that Random Forest is the top performer, achieving 99.93% accuracy and balanced metrics across all criteria evaluated. These findings will significantly enhance web application security by providing reliable defenses against evolving XSS threats.

Aging trajectory prediction of lithium-ion batteries based on mechanical-electrical features via nonlinear autoregressive and regression neural networks

The prediction of the lifespan of lithium-ion batteries (LIBs) is crucial for gaining early insights into battery degradation, enabling the optimization of battery usage and management strategies. This work assesses the expansion and aging characteristics of LIBs under various operating conditions and develops a combined machine-learning method to predict the aging trajectory of LIBs based on mechanical and electrical features. To avoid errors caused by secondary data processing, five mechanical-electrical features are directly extracted from the stress and electrochemical characteristic curves to reflect the battery aging state. A nonlinear autoregressive (NAR) neural network is adopted to predict the change trend of aging features (AFs) using historical data. Subsequently, the nonlinear regression (NR) neural network is employed for aging trajectory prediction using the aging feature dataset, which consists of both historical and predicted data. The experimental results demonstrate that the proposed method can ensure that the mean absolute error (MAE) in aging trajectory prediction is within 0.25 %. The method comprehensively analyzes and utilizes the mechanical-electrical properties of LIBs to enhance the predictive accuracy, expanding the scope of the predictive model in practical applications.

Introducing the unit Zeghdoudi distribution as a novel statistical model for analyzing proportional data

Unit distributions are essential in statistical modeling, providing a robust framework for understanding variables constrained within the unit interval [0,1]. This interval is crucial in public health, environmental studies, engineering, and finance, where measurements often represent proportions, probabilities, or rates. Despite numerous unit distributions derived by transforming existing distributions, there remains a pressing need for new distributions that can accommodate the unique characteristics of diverse datasets. In this study, we introduce the unit Zeghdoudi distribution (UZD), a novel transformation of the Zeghdoudi distribution (ZD). This new distribution retains the simplicity of the original ZD while offering enhanced flexibility and precision in modeling data confined to the unit interval. The UZD exhibits several noteworthy features, including both right and left-skewed probability density functions, a closed-form quantile function, easily defined moments, and manageable entropies. Using sixteen classical methods for parameter estimation, supported by a comprehensive simulation study, we demonstrate the efficiency and reliability of the UZD. Finally, the application of the UZD to three real-world proportional datasets related to (COVID-19, environmental, and radiation datasets) underscores its effectiveness and demonstrates its superiority over several established models.

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.

Pneumonia Detection from Chest X-Ray Images Using Deep Learning and Transfer Learning for Imbalanced Datasets.

Pneumonia remains a significant global health challenge, necessitating timely and accurate diagnosis for effective treatment. In recent years, deep learning techniques have emerged as powerful tools for automating pneumonia detection from chest X-ray images. This paper provides a comprehensive investigation into the application of deep learning for pneumonia detection, with an emphasis on overcoming the challenges posed by imbalanced datasets. The study evaluates the performance of various deep learning architectures, including visual geometry group (VGG), residual networks (ResNet), and Vision Transformers (ViT) along with strategies to mitigate the impact of imbalanced dataset, on publicly available datasets such as the Chest X-Ray Images (Pneumonia) dataset, BRAX dataset, and CheXpert dataset. Additionally, transfer learning from pre-trained models, such as ImageNet, is investigated to leverage prior knowledge for improved performance on pneumonia detection tasks. Our investigation extends to zero-shot and few-shot learning experiments on different geographical regions. The study also explores semi-supervised learning methods, including the Mean Teacher algorithm, to utilize unlabeled data effectively. Experimental results demonstrate the efficacy of transfer learning, data augmentation, and balanced weight in addressing imbalanced datasets, leading to improved accuracy and performance in pneumonia detection. Our findings emphasize the importance of selecting appropriate strategies based on dataset characteristics, with semi-supervised learning showing particular promise in leveraging unlabeled data. The findings highlight the potential of deep learning techniques in revolutionizing pneumonia diagnosis and treatment, paving the way for more efficient and accurate clinical workflows in the future.