Time Data Research Articles

A Study on Partial Discharge Fault Identification in GIS Based on Swin Transformer-AFPN-LSTM Architecture

Aiming at the problem of manual feature extraction and insufficient mining of feature information for partial discharge pattern recognition under different insulation faults in GIS, a deep learning model based on phase and timing features with Swin Transformer-AFPN-LSTM architecture is proposed. Firstly, a GIS insulation fault simulation experimental platform is constructed, and the PRPD phase data and TRPD timing data under different faults are obtained; secondly, the TRPD timing data are converted into MTF; then the PRPD phase data and MTF timing data are input into the Swin Transformer-AFPN-LSTM model and other deep learning models for performance comparison. The experimental results show that the Swin Transformer-AFPN-LSTM model improves the performance by 14.09–21.23% compared with the traditional CNN model and LSTM model. Moreover, using this model to extract phase features and timing features simultaneously improves the accuracy by 10.67% and 8.66%, respectively, compared with single feature extraction, and the overall accuracy reaches 98.82%, which provides a new idea for GIS insulation fault identification.

Application with Deep Learning Framework for Early Prediction of Diabetic Retinopathy

Diabetic retinopathy (DR) is a severe microvascular complication of diabetes that affects the eyes, leading to progressive damage to the retina and potential vision loss. Timely intervention and detection are crucial for preventing irreversible damage. With the advancement of technology, deep learning (DL) has emerged as a powerful tool in medical diagnostics, offering a promising solution for the early prediction of DR. This study compares four convolutional neural network architectures, DenseNet201, ResNet50, VGG19, and MobileNetV2, for predicting DR. The evaluation is based on both accuracy and training time data. MobileNetV2 outperforms other models, with a validation accuracy of 78.22%, and ResNet50 has the shortest training time (15.37 s). These findings emphasize the trade-off between model accuracy and computational efficiency, stressing MobileNetV2’s potential applicability for DR prediction due to its balance of high accuracy and a reasonable training time. Performing a 5-fold cross-validation with 100 repetitions, the ensemble of MobileNetV2 and a Graph Convolution Network exhibits a validation accuracy of 82.5%, significantly outperforming MobileNetV2 alone, which shows a 5-fold validation accuracy of 77.4%. This superior performance is further validated by the area under the receiver operating characteristic curve (ROC) metric, demonstrating the enhanced capability of the ensemble method in accurately detecting diabetic retinopathy. This suggests its competence in effectively classifying data and highlights its robustness across multiple validation scenarios. Moreover, the proposed clustering approach can find damaged locations in the retina using the developed Isolate Regions of Interest method, which achieves almost a 90% accuracy. These findings are useful for researchers and healthcare practitioners looking to investigate efficient and effective powerful models for predictive analytics to diagnose diabetic retinopathy.

The Oregis Dashboard: Web-based Ophthalmic Research Benchmarking in Germany.

The oregis registry operated by the German Ophthalmological Society (DOG) serves as a central digital platform for collecting and analysing ophthalmological healthcare data in Germany. The aim of oregis is to provide a comprehensive picture of the current healthcare situation and promote healthcare research using real-world-data from inpatient and outpatient facilities. Since its launch in 2020, oregis has continuously expanded its database to enable scientific analyses on a wide range of topics. This paper presents the new dashboard feature allowing participating centres to compare their own patient data with aggregated data in real time while also covering privacy aspects of the system. The oregis steering committee opted to implement the oregis dashboard based on existing medical registries. First, forty-nine national and international registries were analysed to identify common features such as benchmarking functionality. This was followed by technical realisation and implementation. Each centre connected to oregis has secure access to the oregis dashboard, which displays key indicators for patients, diagnoses, and treatments. Dynamic filtering options allow targeted data analysis, comparing each centre's results anonymously with aggregated data from other centres. The dashboard uses Apache Superset data visualisation software [! etwas informativer]. Data is synchronised using an integrated oregis Konnektor module to anonymise patient data according to defined standards. Extensive security and privacy measures ensure data security, including server-side encryption, transport encryption, and two-factor authentication. The dashboard is part of a comprehensive privacy policy developed and validated for oregis by privacy experts. The number of centres connected to oregis is still growing. The new web-based dashboard allows flexible analysis of data and comparison with other centres without allowing conclusions to be drawn about any individual care centre; this ensures data privacy and independence in each centre. More features can be added to provide a more thorough overview of eye care in Germany as oregis grows.

Integrated Bayesian Approach for Station Location and Orientation in Marine Active-Source Seismic Studies

In marine active-source seismic studies wherein ocean bottom seismographs are deployed via free-fall from a ship, accurately determining the location and orientation of the instrument poses challenges, especially when relying solely on the travel times of water-borne waves from air gun shots to the station, particularly with a limited distribution of shots. In this study we develop and investigate the utility of a novel methodology that integrates both travel time and waveform polarization data of acoustic waves to enhance the precision of station location. Utilizing polarization data has the potential to significantly reduce location uncertainty, particularly in the across-line direction, and provides a valuable estimate of the orientation of the horizontal channels. Employing Bayesian inference allows for mapping the relative likelihood of station position, considering tradeoffs due to shot geometry and several embedded assumptions. We explore the utility of particle motion polarization information to enhance instrument location and discuss various factors that come into play when employing polarization measurements.

OAH-Net: a deep neural network for efficient and robust hologram reconstruction for off-axis digital holographic microscopy

Off-axis digital holographic microscopy is a high-throughput, label-free imaging technology that provides three-dimensional, high-resolution information about samples, which is particularly useful in large-scale cellular imaging. However, the hologram reconstruction process poses a significant bottleneck for timely data analysis. To address this challenge, we propose a novel reconstruction approach that integrates deep learning with the physical principles of off-axis holography. We initialized part of the network weights based on the physical principle and then fine-tuned them via supersized learning. Our off-axis hologram network (OAH-Net) retrieves phase and amplitude images with errors that fall within the measurement error range attributable to hardware, and its reconstruction speed significantly surpasses the microscope’s acquisition rate. Crucially, OAH-Net, trained and validated on diluted whole blood samples, demonstrates remarkable external generalization capabilities on unseen samples with distinct patterns. Additionally, it can be seamlessly integrated with other models for downstream tasks, enabling end-to-end real-time hologram analysis. This capability further expands off-axis holography’s applications in both biological and medical studies.

Standardization of high-energy laser impact tests with respect to temperature, reflection and perforation time

Abstract High-energy laser impact tests on identical materials and irradiation parameters often show deviating results between different testing institutes. This particularly affects the specular component of the reflection and the temperatures. Even the supposedly easy-to-determine perforation time sometimes shows large variations between different test institutes. For a reliable and reproducible determination of test results, a standardization of the experimental setups is necessary. Test results using two different reference materials are presented. These tests demonstrate that standardizing experimental setups and evaluation methods significantly reduces the deviations of the experimental results. The presented multi-institutional approach aims at gaining data of laser reflection, temperature, and perforation time of laser impacted samples, and at developing an optimized experimental setup based on different existing laboratory equipment and experimental approaches.

Drying Time, Energy and Exergy Efficiency Prediction of Corn (Zea mays L.) at a Convective-Infrared-Rotary Dryer: Approach by an Artificial Neural Network

Energy consumption in the drying industry has made drying an energy-intensive operation. In this study, the drying time, quality properties (color, shrinkage, water activity and rehydration ratio), specific energy consumption (S.E.C), thermal, energy and exergy efficiency of corn drying using a hybrid dryer convective-infrared-rotary (CV-IR-D) were analyzed. In addition, the energy parameters and exergy efficiency of corn were predicted using the artificial neural network (ANN) technique. The experiments were conducted at three rotary rotation speeds of 4, 8 and 12 rpm, drying temperatures of 45, 55 and 65 °C, and infrared power of 0.25, 0.5 and 0.75 kW. By increasing drying temperature, infrared power and rotary rotation speed, the drying time, S.E.C and water activity decreased while the Deff, energy, thermal and exergy efficiency increased. In addition, the highest values of rehydration ratio and redness (a*) and the lowest values of shrinkage, brightness (L*), yellowness (b*) and color changes (ΔE) were obtained at an infrared power of 0.5 kW, air temperature of 55 °C and rotation speed of 8 rpm. The range of changes in S.E.C, energy, thermal and exergy efficiency during the corn drying process was 5.05–28.15 MJ/kg, 3.26–29.29%, 5.5–32.33% and 21.22–55.35%. The prediction results using ANNs showed that the R for the drying time, S.E.C, thermal, energy and exergy data were 0.9938, 0.9906, 0.9965, 0.9874 and 0.9893, respectively, indicating a successful prediction.

Estimation and Variable Selection for Interval-Censored Failure Time Data with Random Change Point and Application to Breast Cancer Study

Motivated by a breast cancer study, we consider regression analysis of interval-censored failure time data in the presence of a random change point. Although a great deal of literature on interval-censored data has been established, there does not seem to exist an established method that can allow for the existence of random change points. Such data can occur in, for example, clinical trials where the risk of a disease may dramatically change when some biological indexes of the human body exceed certain thresholds. To fill the gap, we will first consider regression analysis of such data under a class of linear transformation models and provide a sieve maximum likelihood estimation procedure. Then a penalized method is proposed for simultaneous estimation and variable selection, and the asymptotic properties of the proposed method are established. An extensive simulation study is conducted and indicates that the proposed methods work well in practical situations. The approaches are applied to the real data from the breast cancer study mentioned above. Supplementary materials for this article are available online, including a standardized description of the materials available for reproducing the work.

Analysis of Fentanyl and Fentanyl Analogs Using Atmospheric Pressure Chemical Ionization Gas Chromatography-Mass Spectrometry (APCI-GC-MS).

Illicit fentanyl and fentanyl analogs are a growing concern in the United States as opioid related deaths rise. Given that fentanyl analogs are readily obtained by modifying the structure of fentanyl, illicit fentanyl analogs appearing on the black market often contain similar structures, making analogue differentiation and identification difficult. Thus, obtaining both precursor and product ion data during analysis is becoming increasingly valuable in fentanyl analog characterization. In this paper, we provide GC column retention time, precursor, and product ion mass spectrum data for 74 fentanyl analogs that were analyzed using atmospheric pressure chemical ionization-gas chromatography-mass spectrometry (APCI-GC-MS) utilizing a triple quadrupole mass analyzer. During analysis, precursor ions underwent collision induced dissociation (CID) by increasing the collision energy (10, 20, 30, 40, and 50 V) throughout a single run. Data reveal that APCI readily produces product ions of the piperidine and N-alkyl chain but rarely provides data on the acyl group. Furthermore, fentanyl analogs with greater substitution at the N-alkyl chain demonstrate a greater preference for dissociation at the N-αC and αC-βC bond, while greater substitution at the amide group leads to fragmentation at the N-C4 bond.

SeqInfo-SAWGAN-GP: Adaptive feature extraction from vibration time data under variable operating conditions for imbalanced bearing fault diagnosis

Current generative adversarial network (GAN)-based methods are widely used to address data imbalance in intelligent fault diagnosis of bearings, a problem often caused by the scarcity of faulty data in real-world scenarios. However, these methods still face challenges, including high model training difficulty and poor quality under variable operating conditions. To address this, sequence information incorporated self-attention Wasserstein generative adversarial network with gradient penalty (SeqInfo-SAWGAN-GP) is proposed. Firstly, a self-attention block is integrated to extract significant features from limited time data. Then, SeqInfo block is introduced to enhance the quality of synthetic time data and manage the model to generate data across different operating conditions. Additionally, healthy vibration signals corresponding to specific operating conditions are input into the generator to further expedite the generation of synthetic fault data. Finally, the superiority of the proposed method is validated on two bearing datasets. The results indicate that the proposed method achieves accuracy of 98.5% and 95% using only a few samples, marking improvements of 4% and 45% over the original imbalanced dataset on two bearing datasets under variable operating conditions, respectively, and outperforming state-of-the-art methods.

A systematic review of machine learning and deep learning techniques for anomaly detection in data mining

The growing use of the internet has increased the threat of cyberattacks. Anomaly detection systems are vital for protecting networks by spotting irregular activities. Various studies investigated anomaly detection techniques without a systematic approach. So far, the existing reviews mainly concerned time series and data streaming methods almost neglected the growing interest in graph-based data mining techniques which are vital in social networks, finance, and IoT domains. Following PRISMA guidelines, this systematic review examines anomaly detection methods applied to time series, data streaming, and graph-based data from 2018 to 2023. A total of 34 papers were selected from four key databases ScienceDirect, Scopus, Web of Science, and IEEE. In addition, this review addressed several issues with existing techniques including low scalability, explainability, and interpretability for real-time anomaly detection systems. In modern applications where data structures are complex, and processing requirements are high these traditional techniques are insufficient for real-time data processing. Finally, our study demanded more advanced, complex methods to address these evolving challenges.

The empirical analysis of multi-objective hyperparameter optimization in software vulnerability prediction

The security of software has always been a vital concern for software developers. Infringements on software systems can result in significant losses in terms of time and confidential data. Software vulnerabilities are considered to be the gateway for attackers to harm information systems. Hence, it is crucial to build effective software vulnerability prediction models. Machine learning algorithms produce high-performance prediction models but the performance of these models is affected by the hyperparameter settings of machine learning methods and imbalanced datasets. The current study aims to find out the effect of single- and multi-objective hyperparameter optimization on software vulnerability prediction models. The paper has proposed an experimental methodology that considers both optimizations and uses eight machine learning methods on PHP open-source public datasets (Drupal, Moodle, and PHPMyAdmin). The experimental results show that after applying multi-objective hyperparameter optimization, the highest AUC achieved is 0.9774 and the F1-Score attained is 0.9222, which is better than the benchmark studies. Random Forest has performed satisfactorily for all three datasets. The comparative analysis of single and multi-objective HPO is performed using post hoc Tukey’s HSD test. Furthermore, the effectiveness of the resampling technique ‘SMOTE’ is observed.

Interplay of aging and practice in conflict processing: A big-data diffusion-model analysis.

We are continually required to exercise cognitive control in order to separate relevant and irrelevant information. Previous studies have produced mixed results as to whether cognitive control declines across adulthood and improves with practice. Moreover, little is known about the influences of aging and practice on the automatic and controlled processes underlying performance in conflict situations. This calls for analyses of extensive data using process models for conflict tasks, akin to earlier drift-diffusion model analyses of performance in cognitive nonconflict tasks. Thus, to understand how aging and practice influence cognitive control at the process level, we analyzed a large-scale data set (1,800 participants aged 21-80 years completing 60 blocklike online games of an arrow-based Eriksen flanker task). At the coarse-grained level of mean response times, the congruency effect increased across adulthood and decreased with practice following an initial increase. The finer-grained distributional response time and error rate data were closely fitted by the diffusion model for conflict tasks, which captures the dynamic interplay of automatic and controlled processing. Best-fitting parameter values revealed multiple, partially counteracting influences of aging and practice: Aging across adulthood slowed down both controlled and automatic processing (besides slowing down nondecisional processes and increasing decision caution). By contrast, practice selectively speeded up controlled processing (besides speeding up nondecisional processes and decreasing decision caution). Taken together, these findings suggest that aging and practice primarily alter the speed of controlled (aging and practice) and automatic processing (aging), rather than causing inhibitory adjustments in the strength of automatic processing. (PsycInfo Database Record (c) 2025 APA, all rights reserved).

Applying the LATER model to reaction time data: an open-source toolkit.

Analyzing reaction time distributions can provide insights into decision-making processes in the brain. The Linear Approach to Threshold with Ergodic Rate (LATER) model is arguably the simplest model for predicting reaction time distributions and can summarize distributions with as few as two free parameters. However, the coordinates for visualizing and fitting distributions with LATER ("reciprobit" space) are irregular, making the application of this simple model inaccessible to those without a programming background. Here we describe an open-source R package, LATERmodel, that allows for easy visualization of reaction time distributions, along with fitting of these with the LATER model. Using canonical data from the literature, we show that our tool replicates key features from previous LATER analysis tools, while also providing more robust fitting procedures and a more flexible method for fitting subpopulations of very rapid, early responses. Although all features of LATERmodel can be used directly in the statistical programming language R, key features are also available through a RShiny graphical user interface to allow researchers without programming background to apply the LATER model to their reaction time data.NEW & NOTEWORTHY Analyzing reaction time distributions provides a powerful tool for investigating decision-making processes. Here we describe an open-source toolbox to allow the Linear Approach to Threshold with Ergodic Rate (LATER) model, the simplest principled model linking reaction times and decisions, to be applied to empirical reaction time data, including by clinicians and scientists without any programming experience.

Combining choice and response time data to analyse the ride-acceptance behavior of ride-sourcing drivers

Combining choice and response time data to analyse the ride-acceptance behavior of ride-sourcing drivers

Approaches for GC-HRMS Screening of Organic Microcontaminants: GC-APCI-IMS-QTOF versus GC-EI-QOrbitrap.

This study explores the capabilities of GC-APCI-IMS-QTOF MS and GC-EI-QOrbitrap MS in screening applications and different strategies for wide-scope screening of organic microcontaminants using target suspect and nontarget approaches. On one side, GC-APCI-IMS-QTOF MS excels at preserving molecular information and adds ion mobility separation, facilitating screening through the list of componentized features containing accurate mass, retention time, CCS, and fragmentation data. On the other side, the extensive and robust fragmentation of GC-EI-QOrbitrap MS allows the application of different strategies for target and nontarget approaches using the NIST library spectra. Our findings revealed that GC-EI-QOrbitrap MS is more sensitive in target approaches. Automated workflows for suspect screening in GC-APCI-IMS-QTOF MS minimize false annotations but face challenges with false negatives due to in-source fragmentation and limitations when using in silico fragmentation tools. Conversely, a nontarget approach in GC-EI-QOrbitrap MS can reliably identify unknowns but results in more false annotations in complex matrices. This work highlights the strengths and limitations of each system and guides for their optimal application for wide-scope screening in environmental and food safety applications.

Implementation of Tracking System Based on JT701 Smart Lock Technology on Goods Delivery Service Vehicles

The goods delivery service industry has emerged as a strategic business both during and after the pandemic. The shift towards online shopping, which became a norm during the pandemic, let’s continue post-pandemic, making it a promising sector for companies involved in goods delivery. However, the industry faces challenges, particularly related to security during shipment. One common issue is the inability to track vehicle positions, leading to unauthorized loading and unloading activities at incorrect locations, causing losses for business owners. Conventional mechanical locks are vulnerable to theft, duplication, key loss, and vandalism, making them inadequate for securing cargo. To address these issues, Smart Lock technology has been developed, offering enhanced security through modern, integrated features. This technology eliminates the need for physical keys, reducing the risk of loss, and allows for remote access management. Additionally, it enables cargo owners to monitor the vehicle's position and control the lock remotely. Given these advancements, there is a need to explore Tracking Systems for delivery vehicles, incorporating automatic locks controlled via Internet of Things (IoT) technology. Based on the measurement, the tracking system tool testing was carried out at 5 different locations. The distances are 2.9 km, 6.4 km, 3 km, 7.1 km, 4.7 km from campus. The GPS signal is sent to the cloud server Cloud of Jointech. Data transmission uses the GSM network. There is a delay difference of 2-3 seconds from the GPS time signal compared to the time data displayed on the website. The data provided from the GPS signal is updated every 20 seconds. Based on the 5 test locations that have been set, the on/off lock test has been successfully carried out.

Crafting a comprehensive strategy to prevent fraud and enable recovery in mobile insurance in the USA

In the U.S., the rapid expansion of mobile insurance has also created an environment of opportunity and challenge around fraud prevention and recovery. With the ever-increasing sector, fraud in the market can be fought using technological advancements like artificial intelligence (AI), machine learning and blockchain; these help in real time data analysis, pattern recognition and secure transactions. These innovations must however, always be perfected to remain one step ahead of the game and always evolving in fraudulent tactics. This paper discusses the important obstacles facing mobile insurance; namely emerging fraud risks, the necessity for comprehensive regulatory frameworks and lack of consumer awareness. The gaps in current fraud prevention tactics are uncovered by the research and the necessary regulatory updates to address digital identity and mobile platforms are highlighted. This highlights the importance for insurers, lawmakers and consumers to develop a collaborative way to gain trust and foster effective fraud detection and recovery programs. The paper also discusses the limitations of existing fraud detection models, and analyzes the opportunities for new legislation and enforcement mechanisms for reducing the risk of fraud. Also, awareness of consumer education as a key fraud prevention tool is considered important, in that educated consumers would be able to avoid fraudulent schemes. It also highlights the importance of additional research in promising technologies to apply to mobile insurance fraud prevention. Through all of this research, insights are provided into the current state of mobile insurance, challenges of fraud prevention, and actionable recommendations for industry stakeholders that will improve security and help create a more trusting mobile insurance ecosystem.

Building a .NET and AI-Powered Collaborative Platform to Strengthen Inter-Entity Relationships and Foster US Economic Growth

In this study, we take stock of how AI powered collaborative platforms, utilizing .NET technologies can be integrated to handle several critical inter-entity collaboration problems and their potential in boosting economic growth of US. Increasingly, demand in business, finance and government organizations is mounting on efficient communication and data sharing, and AI and .NET frameworks are offering strong solutions at hand. In particular, the research shows that these technologies can enhance collaboration in terms of scalability, system reliability, data security and processing of large volumes of real time data. For organizations, this implementation of AI driven tools such as predictive analytics, machine learning and real time monitoring can be used to make better decisions, boost productivity and build better inter organizational relationships. The challenge of integrating AI and .NET frameworks is evident though they have the potential. Systems are not compatible, infrastructure for legacy applications is a limitation, and it is a complex matter to make sure real time data can be exchanged between dissimilar platforms. Secondly, data privacy and security risks, and scalability problems in large—scale environment are critical areas of concern. This study looks in detail at how these challenges are being tackled, and what further refinements are required to enhance the performance and reliability of collaborative platforms. The results of this research provide important implications for business, policy makers and financial institutions wanting to develop economies by using AI and .NET technologies. In addition, the paper highlights research gaps and emerging issues related to AI powered collaboration, which may drive future innovation, technological development and policy reforms. Finally, the research hopes to guide the creation of more efficient, scalable, and secure collaborative platforms that can support sustainable U.S. economy growth.

Design of a Multi-Node Data Acquisition System for Logging-While-Drilling Acoustic Logging Instruments Based on FPGA

The logging-while-drilling (LWD) acoustic logging instrument is pivotal in unconventional oil and gas exploration, and in providing real-time assessments of subsurface formations. The acquisition system, a core component of the LWD acoustic logging suite, is tasked with capturing, transmitting, and processing acoustic signals from the formation, which directly affects the accuracy and timeliness of the logging data. Recognizing the constraints of current LWD acquisition systems, including limited data collection capabilities and inadequate precision, this study introduces an FPGA-based multi-node data acquisition system for LWD acoustic logging. This system increases sampling density and data accuracy, leading to a more comprehensive collection of formation information. The multi-node acquisition system is composed primarily of a main control circuit board and several acquisition circuit boards, all connected via an RS485 bus. The Field-Programmable Gate Array (FPGA) is utilized to develop the acquisition circuit board’s firmware, offering adjustable control over parameters, such as the AD7380’s operational mode, sampling rate, and depth, facilitating real-time and concurrent acquisition and storage of formation acoustic signals. The main control board communicates with the acquisition boards via the RS485 bus, issuing commands to enable autonomous data collection and transfer from each board, thus enhancing the system’s reliability and scalability. Experimental results confirm the system’s capacity to efficiently capture waveform signals and upload them in real-time, underscoring its dependability and timeliness. The findings suggest that the system is capable of high-speed, real-time acquisition and processing of acoustic signals, offering robust technical support for the continued application of LWD acoustic logging instruments.