Time Prediction Model Research Articles

Corrigendum to “Prediction model for equivalent exposure time in indoor dry-wet cycling tests for concrete in marine tidal zones” [Constr. Build. Mater. 451 (2024) 138778

Corrigendum to “Prediction model for equivalent exposure time in indoor dry-wet cycling tests for concrete in marine tidal zones” [Constr. Build. Mater. 451 (2024) 138778

An OD time prediction model based on adaptive graph embedding

An OD time prediction model based on adaptive graph embedding

Additional Taxi-Out Time Prediction for Flights at Busy Airports by Fusing Flow Control Information

The taxi-out time of an airport scene can be categorized into the unimpeded taxi-out time and the additional taxi-out time. Usually, additional taxi-out time is used as a key index to monitor taxi-out performance, and its accurate prediction plays an important role in optimizing the allocation of time slots at an airport and improving scene operation efficiency. Taking Shanghai Pudong International Airport as the research object, we first analyze its layout and construct the origin–destination pairs (ODPs) based on the stand groups and runways. Then, we develop a multiple linear regression model based on the arrival and departure flows to calculate the unimpeded taxi-out times for all ODPs. The actual taxi-out time is then subtracted from the unimpeded taxi-out time to obtain the historical additional taxi-out time of each flight. We propose three new flow features related to the structure: the corridor departure flow, the corridor arrival flow, and the departure flow proportion of ODPs, based on which we construct a dataset for training the prediction model. We then propose an additional taxi-out time prediction model based on the nutcracker optimization algorithm (NOA) and XGBoost and run comparison experiments on the operation data of our target airport. The results show that the optimized prediction model we proposed has the best performance compared with the traditional XGBoost model and other commonly used prediction models, and the proposed structure-related features have high correlations with additional taxi-out time.

Prediction model for equivalent exposure time in indoor dry-wet cycling tests for concrete in marine tidal zones

Chloride ion erosion poses a major threat to the durability of concrete structures in coastal environments, particularly in tidal zones with periodic seawater exposure. This study developed a predictive model based on Fick's second law, utilizing data from over 20 coastal structures across six countries. The model, grounded in the principle of equivalent free chloride ion concentration at a certain depth after a given exposure time, estimates the number of dry-wet cycles required in accelerated laboratory testing to simulate real-world conditions. It has been validated using data from actual marine structures. Key environmental factors such as temperature and humidity, along with regional climate differences, are incorporated to predict the relative deterioration rates of concrete in various marine environments. Additionally, an accelerated indoor dry-wet cycle test was designed based on the theory of environmental similarity, tracking the performance evolution of concrete. The model was verified through experiments, showing that 30 indoor dry-wet cycles correspond to 79 years of exposure in a coastal environment. The physical property changes observed in the accelerated tests further confirmed the model’s reliability and accuracy. This research provides critical insights into the impact of environmental factors on concrete degradation and offers a scientific basis for the application of accelerated laboratory testing.

Rutting prediction model of asphalt pavement based on RIOHTrack full-scale ring road

The research and prediction of rutting mechanism of asphalt pavement has always been the main difficult problem for road designers all over the world, and scholars have done a lot of research on how to prevent and control rutting of asphalt pavement. Based on the measured loading data of 80 million axis times of Beijing RIOHTrack full-scale ring road, rutting evolution of asphalt pavement with various structure types is studied in this paper and proposes the Kou&Cao rutting prediction model for the first time. Compared with other mature explicit rutting prediction models, this model has the characteristics of high prediction accuracy, strong robustness. At the same time, the model can provide a basis for predicting the inflection point of rutting evolution on asphalt pavement. The research results of this model are effective, which completes the effective analysis of the rutting evolution and development law of asphalt pavement, and provides an important reference for the future road design specification research and road construction.

A Disturbance-Driven Textual Model for Train Running Time Prediction on High-Speed Railways

A Disturbance-Driven Textual Model for Train Running Time Prediction on High-Speed Railways

HYBRID ARCHITECTURE WITH IMPROVED SCORE LEVEL FUSION FOR PATIENT WAITING TIME PREDICTION

Hospitals are overcrowded in proportion to the massive increase in population, making it difficult for hospitals to manage or arrange or shorten patient wait time during medical care. However, it is difficult to overstate the significance of patient waiting time management and predictability. In most of the hospitals, the clinical workflow is defined and driven by the patient queues. Predicting the incoming patients and waiting times has thus emerged as one of the most crucial clinical management techniques. Our proposed work aims to address the limitations of current waiting time prediction models by providing a more sophisticated, data-driven approach that can adapt to the complexities of healthcare environments. This work focuses on developing a new model for predicting the patient’s waiting time. The primary aim of this research is to create a reliable predictive model capable of accurately anticipating patient waiting times. This innovation aims to enhance both patient satisfaction and operational efficiency within healthcare settings. By offering healthcare administrators a tool to better manage patient flow, allocate resources efficiently, and minimize waiting times, this model seeks to optimize overall healthcare service delivery. The initial process carried out is pre-processing which includes data acquisition and improved [Formula: see text]-score normalization. Subsequently, entropy features, correlation features, and improved mutual information (MI) features are extracted. Then, prediction is carried out using hybrid classifier (HC) that involves a convolutional neural network (CNN) and bidirectional gated recurrent unit (Bi-GRU). Finally, improved score level fusion (ISLF) is introduced to get absolute outcomes on WTP. The analysis outcomes on varied error metrics show the efficacy of the proposed WTP model.

(S)-2,5-dioxopyrrolidin-1-yl-1-(4,6-dimethoxy-1,3,5-triazin-2-yl)pyrrolidine-2-carboxylate derivatization combined with UPLC/ESI-MS/MS as a powerful method for dipeptide determination in foods: An application example in fermented cocoa beans

Dipeptides play a crucial role in food flavor, as well as in physiological and nutritional functions. However, as comprehensive methods for determining dipeptides are still under development, the dipeptide composition of foods remains largely unexplored. This study proposed a novel analytical method for dipeptides by applying an amino acid determination technique we previously developed: (S)-2,5-dioxopyrrolidin-1-yl-1-(4,6-dimethoxy-1,3,5-triazin-2-yl)pyrrolidine-2-carboxylate (DMT-(S)-Pro-OSu) derivatization coupled with ultra-performance liquid chromatography–electrospray ionization tandem mass spectrometry (UPLC/ESI-MS/MS). The DMT-(S)-Pro-OSu derivatization of dipeptides improved both separation in LC and detection sensitivity in MS, allowing for the optimized acquisition of m/z and elution times for 313 available dipeptides, and enabling the absolute quantification of 126 dipeptides in a simultaneous analysis. Subsequent analysis of fermented cocoa beans detected and quantified 42 dipeptides, revealing distinct dipeptide profiles depending on the origin of the beans. Additionally, by conducting in silico analysis of the elution times for 313 dipeptides, a machine learning model was constructed that accurately predicts dipeptide elution times. The DMT-(S)-Pro-OSu-UPLC/MS/MS method and the elution time prediction model developed in this study are expected to contribute to the understanding of dipeptide composition in foods.

Analysis of Factors Influencing Wait Times at Disneyland Resort

This paper aims to enhance the accuracy of wait time predictions and optimize park management at Disneyland, a popular leisure destination facing long wait times for attractions. These wait times adversely affect visitors' immediate experiences and overall evaluations, influencing their willingness to return. These wait times adversely affect visitors' immediate experiences and overall evaluations, influencing their willingness to return. The study employs advanced data analysis and machine learning algorithms to develop high-precision wait time prediction models. Utilizing big data techniques, the research integrates multi-dimensional data, including weather, visitor flow, and historical wait times, to optimize resource allocation and improve operational efficiency. optimize resource allocation and improve operational efficiency and visitor satisfaction. The study further explores the application of these models in other domains such as healthcare, healthcare, and healthcare services. The study further explores the application of these models in other domains such as healthcare and transportation, demonstrating the potential for cross-disciplinary technology transfer to enhance overall service quality and efficiency. The findings underscore the significant impact of big data behavior analysis in theme park management, contributing to the findings underscore the significant impact of big data behavior analysis in theme park management, contributing to increased visitor satisfaction, enhanced competitiveness, and long-term sustainability of the park.

Temporal dynamics changes in the vaginal fluid microbiome: Implications for body fluid identification and estimating time since deposition (TsD) for forensics

Vaginal fluid analysis plays a crucial role in sexual assault investigations. However, vaginal fluid found at crime scenes is usually subject to a certain duration of exposure. This study thus aimed to assess the influence of different durations of exposure to indoor environment on the vaginal microbiota. The 16S rDNA high-throughput sequencing was used on vaginal fluid samples exposed for short-term (30 days) and long-term (240 days), respectively. Despite potential contamination from environmental microorganisms, particularly following long-term exposure, the results indicated that the vaginal microbiota after exposure was still dominated by Lactobacillus. Both in short-term and long-term exposure involving vaginal fluid, there were clusters with time-dependent characteristics, wherein the relative abundances of associated microbial genera showed a trend of increasing or decreasing over time. In addition, each bodily fluid presented with a unique array of dominant bacterial genera, enabling the differentiation of exposed vaginal fluid samples from other bodily fluids (semen, skin, saliva, feces) with a remarkable 98.75 % accuracy rate. Furthermore, the mean absolute error achieved by the long-term deposition time prediction model was 13.54 days. The mean absolute error for the short-term deposition time prediction model was notably lower, reaching just 2.05 days. In summary, this study investigates the variations in microbial communities within vaginal fluid subjected to different indoor exposure durations and explores their potential in body fluid identification and estimating the time since deposition, thereby contributing valuable supporting evidence in forensic investigations.

Mechanisms of water layer thickness and ullage height on crude oil boilover: a theoretical model coupling the effects of multiple physical fields

Boilover is one of the most destructive tank fire scenarios. A series of experiments were conducted using eight different depths of oil pans (ranging from internal depths of 4 to 20 cm) to vary the water layer thickness and ullage height. The results indicate that the water layer effectively cools the sidewalls, reduces the burning rate, inhibits the development of hot zones, and delays the onset of boilover in small and medium-scale experiments. Conversely, the ullage height affects the burning rate, formation of hot zones, intensity of the boilover, and boilover onset time. Utilizing experimental data and thermodynamic analysis, both water layer thickness and fuel layer thickness were considered as variables to predict sidewall temperature at the fuel surface. These results were then introduced into the burning rate prediction model. A prediction model for the boilover onset time was also developed using the water layer thickness as a variable, and a thermodynamic analysis revealed the existence of a limit to the effect of water layer thickness on the boilover onset time. Bubble dynamics was introduced to analyze the boilover process at the oil-water interface, clarifying that the influence of water layer thickness and ullage height on boilover intensity primarily lies in factors such as the degree of superheat at the fuel-water interface. The study’s findings hold significant implications for predicting and assessing fire accidents in storage tanks.

PowerTrain: Fast, generalizable time and power prediction models to optimize DNN training on accelerated edges

Accelerated edge devices, like Nvidia’s Jetson with 1000+ CUDA cores, are increasingly used for DNN training and federated learning, rather than just for inferencing workloads. A unique feature of these compact devices is their fine-grained control over CPU, GPU, memory frequencies, and active CPU cores, which can limit their power envelope in a constrained setting while throttling the compute performance. Given this vast 10k+ parameter space, selecting a power mode for dynamically arriving training workloads to exploit power–performance trade-offs requires costly profiling for each new workload, or is done ad hoc. We propose PowerTrain, a transfer-learning approach to accurately predict the power and time that will be consumed when we train a given DNN workload (model + dataset) using any specified power mode (CPU/GPU/memory frequencies, core-count). It requires a one-time offline profiling of 1000s of power modes for a reference DNN workload on a single Jetson device (Orin AGX) to build Neural Network (NN) based prediction models for time and power. These NN models are subsequently transferred (retrained) for a new DNN workload, or even a different Jetson device, with minimal additional profiling of just 50 power modes to make accurate time and power predictions. These are then used to rapidly construct the Pareto front and select the optimal power mode for the new workload, e.g., to minimize training time while meeting a power limit. PowerTrain’s predictions are robust to new workloads, exhibiting a low MAPE of <6% for power and <15% for time on six new training workloads (MobileNet, YOLO, BERT, LSTM, etc.) for up to 4400 power modes, when transferred from a ResNet reference workload on Orin AGX. It is also resilient when transferred to two entirely new Jetson devices (Xavier AGX and Jetson Orin Nano) with prediction errors of <14.5% and <11%. These outperform baseline predictions by more than 10% and baseline optimizations by up to 45% on time and 88% on power.

A novel query execution time prediction approach based on operator iterate-aware of the execution plan on the graph database

Query execution time prediction is essential for database query optimization tasks, such as query scheduling, progress monitoring, and resource allocation. In the query execution time prediction tasks, the query plan is often used as the modeling object of a prediction model. Although the learning-based prediction models have been proposed to capture plan features, there are two limitations need to be considered more. First, the parent–child dependencies between plan operators can be captured, but the operator’s branch independence cannot be distinguished. Second, each operator’s output row is its following operator input, but the data iterate transfer operations between operators are ignored. In this study, we propose a graph query execution time prediction model containing a plan module, a query module, a plan-query module, and a prediction module to improve prediction effectiveness. Specifically, the plan module is used to capture the data iterate transfer operations and distinguish independent of branch operators; the query module is used to learn features of query terms that have an influence on the composition of operators; the plan-query interaction module is used to learn the logical correlations of plan and query. The experiment on datasets proves the effectiveness of the operator iterate-aware and query-plan interaction method in our proposed graph query execution prediction model.

Novel extended hybrid tool for real time control and practically support decisions to reduce GHG emissions in full scale wastewater treatment plants

In this paper, a novel methodology and extended hybrid model for the real time control, prediction and reduction of direct emissions of greenhouse gases (GHGs) from wastewater treatment plants (WWTPs) is proposed to overcome the lack of long-term data availability in several full-scale case studies. A mechanistic model (MCM) and a machine learning (ML) model are combined to real time control, predict the emissions of nitrous oxide (N2O) and carbon dioxide (CO2) as well as effluent quality (COD – chemical oxygen demand, NH4-N – ammonia, NO3-N - nitrate) in activated sludge method. For methane (CH4), using the MCM model, predictions are performed on the input data (VFA, CODs for aerobic and anaerobic compartments) to the MLM model. Additionally, scenarios were analyzed to assess and reduce the GHGs emissions related to the biological processes. A real WWTP, with a population equivalent (PE) of 125,000, was studied for the validation of the hybrid model. A global sensitivity analysis (GSA) of the MCM and a ML model were implemented to assess GHGs emission mechanisms the biological reactor. Finally, an early warning tool for the prediction of GHGs errors was implemented to assess the accuracy and the reliability of the proposed algorithm. The results could support the wastewater treatment plant operators to evaluate possible mitigation scenarios (MS) that can reduce direct GHG emissions from WWTPs by up to 21%, while maintaining the final quality standard of the treated effluent.

Optimization of garment sewing operation standard minute value prediction using an IPSO-BP neural network

Abstract Standard minute value serves as a pivotal metric guiding the arrangement and balancing of production cycles in clothing production lines, and plays a crucial role in cost pricing and production order arrangement for clothing products. Given the complexity of the garment sewing process, ten influencing factors including fabric weight, fabric thickness, fabric density, stitching length, stitching shapes, cut pieces numbers, notch numbers, sewing technologies, sewing machine, and auxiliary accessories were identified. Upon this foundation, a standard sewing time prediction model, Improved particle swarm optimization - Back-propagation neural network (IPSO-BP), was proposed, focusing on non-quantitative factors. The IPSO-BP model was trained using actual sewing data from a women’s clothing production company. Compared to the unoptimized BP neural network, the IPSO-BP model demonstrated significant advantages in terms of convergence speed and prediction accuracy. Therefore, the IPSO-BP model proposed in this study holds promise for predicting standard sewing hours effectively.

A federated data fusion-based prognostic model for applications with multi-stream incomplete signals

Most prognostic methods require a decent amount of data for model training. In reality, however, the amount of historical data owned by a single organization might be small or not large enough to train a reliable prognostic model. To address this challenge, this article proposes a federated prognostic model that allows multiple users to jointly construct a failure time prediction model using their multi-stream, high-dimensional, and incomplete data while keeping each user’s data local and confidential. The prognostic model first employs multivariate functional principal component analysis to fuse the multi-stream degradation signals. Then, the fused features coupled with the times-to-failure are utilized to build a (log)-location-scale regression model for failure prediction. To estimate parameters using distributed datasets and keep the data privacy of all participants, we propose a new federated algorithm for feature extraction. Numerical studies indicate that the performance of the proposed model is the same as that of classic non-federated prognostic models and is better than that of the models constructed by each user itself.

Multistep Ahead Taxi-Out Time Prediction Using Long Short-Term Memory Networks

For more reliable scheduling and management of air traffic, there is a need for a more accurate prediction of the taxi-out time before aircraft departure from the gate. This paper presents a multistep ahead taxi-out time prediction model that utilizes long short-term memory, considering airport surface congestion before aircraft departure. The comparison results showed that the prediction performance of the proposed model improves the accuracy compared to the three existing regression models and that the prediction error of the proposed model is consistent regardless of the traffic congestion on the airport surface.

Multi-Task Dynamic Spatio-Temporal Graph Attention Network: A Variable Taxi Time Prediction Model for Airport Surface Operation

Variable taxi time prediction is the core of the Airport Collaborative Decision Making (A-CDM) system. An accurate taxi time prediction contributes to enhancing airport operational efficiency, safety and predictability. The deep dynamic spatio-temporal correlation inherent in airport traffic data is critical for taxi time prediction. However, existing machine learning (deep learning) methods have been unable to thoroughly exploit these correlations. To address this issue, we propose a deep learning-based model called the multi-task dynamic spatio-temporal graph attention network (MT-DSTGAN). Our model also predicts future entire airport traffic flow and taxiing segment traffic flow as auxiliary tasks, with the goal of enhancing the accuracy of aircrafts’ taxi time prediction. The proposed MT-DSTGAN model is implemented and assessed through a case study of Beijing Capital International Airport with a real-world dataset. The advantage of the proposed model, which shows better performance in various evaluation metrics, is demonstrated in a comparative study with other baseline works. In summary, the proposed MT-DSTGAN exhibits promising capabilities in perceiving the dynamic changes in the taxiing process of aircraft and demonstrates the ability to capture complex spatio-temporal correlations in airport traffic data.

Using data-driven techniques to predict outpatient ultrasound examination time for the multi-clinic outpatient appointment scheduling problem

Appointment scheduling for hospital patients is becoming increasingly important because it can reduce patient’s waiting time and enhance hospitals’ medical quality. In the case of ultrasound examinations, multiple radiological technologists use multiple rooms to conduct examinations, which involves the sequencing of examination work, time uncertainty, and operational characteristics. To enable hospital managers to effectively evaluate radiological technologists’ effort when carrying out ultrasound examinations, this study uses classification methods in data exploration, including decision tree (DT), logistic regression (LR), and artificial neural network (ANN), to establish a time prediction model for outpatient ultrasound examination. This study found that the DT model (i.e. C4.5) had the highest classification accuracy. After reviewing and removing rules generated from the model with an accuracy below 70%, the overall accuracy of the model was 76.31%. The classification results of the model are a reference for outpatient ultrasound appointment scheduling. The results of the study can effectively predict the examination time needed for ultrasound outpatients requiring different examination positions, thereby fairly allocating the patients to each room for the examination and reducing patient waiting time and radiological technologists’ idle time. Therefore, hospital managers can improve the hospital’s overall medical quality.

Stratified Patient Profiling for Vedolizumab Effectiveness in Crohn's Disease: Identifying Optimal Subgroups for Enhanced Treatment Response in the EVOLVE Study.

This analysis evaluated the relative performance of vedolizumab and anti-tumor necrosis factor alpha (anti-TNFα) agents in subpopulations of biologic therapy-naive patients with Crohn's disease (CD) and assessed whether patients in whom vedolizumab would have a larger treatment effect vs anti-TNFα agents could be identified. Data were from EVOLVE, a real-world, multicountry, retrospective cohort study of patients with inflammatory bowel disease who initiated first-line biologic treatment with vedolizumab (n = 195) or anti-TNFα agents (n = 245). Prediction models for time to clinical remission were developed in vedolizumab- and anti-TNFα-treated patients and used to estimate effect scores, a metric of predicted comparative efficacy, for each patient. Patients were ranked by effect scores and potential subpopulations were investigated. Simplified rules to identify these subpopulations were also developed using classification tree analysis. Among all patients, median time to clinical remission was 7.8months (vedolizumab) and 11.1months (anti-TNFα) (P < 0.05). Among patients in the top 40% of the effect score distribution, the median time to clinical remission was 4.8months (vedolizumab) vs 18.1months (anti-TNFα) (adjusted hazard ratio 2.0, 95% confidence interval 1.3-2.9). A simplified rule for identifying a subpopulation more likely to benefit from vedolizumab was based on having an ongoing CD exacerbation, no prior emergency visits, and non-stricturing disease. Subpopulations of biologic-naive patients with CD in whom vedolizumab appeared to have a larger effect relative to anti-TNFα agents for the outcome of clinical remission were identified. Validation of the identified subpopulations and simplified rules are warranted to confirm these findings. NCT03710486. Graphical Abstract available for this article.