Dynamic Prediction Research Articles

Enhancing Supply Chain Efficiency with Predictive Analytics: A Machine Learning Approach to Dynamic Lead Time Prediction in a Chemical Industry

We develop and employ a four-step methodological approach for predicting the lead time delays in all echelons of a supply chain (SC). The first step of the methodological approach involves a critical synthesis of academic research efforts for identifying the main sources of delays in all echelons of a supply chain. The second step involves the development of questionnaires for validating the findings of the research through workshops with industry stakeholders. The third step involves the development of a suite of machine learning (ML) models, namely, Random Forest Regression, Decision Tree Regression, and Linear Regression. These models were selected based on their prevalence in the recent literature and their ability to handle linear and nonlinear relationships between multiple variables. The final fourth step involves the implementation of the suite of machine learning models in the real case of a Hellenic chemical manufacturing supply chain. The implementation results reveal that Random Forest Regression exhibits the highest predictive accuracy throughout all stages of the supply chain, achieving the lowest Mean Absolute Percentage Errors (MAPE), ranging from 0.5 to 7% in the examined supply chain echelons.

A dynamic strain prediction method for malfunction of sensors in buildings subjected to seismic loads using CWT and CNN.

Structural health monitoring (SHM) is a technique used to evaluate the safety of a structure based on the structural response. In the SHM, sensors may fail to measure the structural response, or data loss can occur during the data transmission and reception. In this regard, this paper proposes a method of predicting the strain response when a sensor installed in a structural member suffers a malfunction. To this end, a convolutional neural network (CNN) is introduced to predict the strain response. Time-history strain data for seismic loads measured in adjacent structural members are set as the input layer of the CNN, and continuous wavelet transform (CWT) results are additionally set in the input layer of the CNN to reflect the dynamic structural behavior during earthquakes in the CNN predictions. Time-history strain data for a member, the target of prediction, are set as the output layer of the CNN. Then, the trained CNN uses the time-history strain data of adjacent members with no sensor failure and the CWT result of the data to predict the strain response of a structural member with sensor failure. A numerical study on the seismic load of a single-span three-story steel moment frame and an experimental study on a single-span three-story reinforced concrete frame were carried out to verify the validity of the proposed method. This study presents a method of applying CWT data to the input layer of the CNN and discusses the effectiveness of CWT data for predicting the nonlinear response of a structural member with damaged sensors.

Vehicle Interactive Dynamic Graph Neural Network-Based Trajectory Prediction for Internet of Vehicles

Vehicle Interactive Dynamic Graph Neural Network-Based Trajectory Prediction for Internet of Vehicles

Refined subsidence monitoring and dynamic prediction in narrow and long mining areas based on InSAR and probabilistic integral method.

Continuous exploitation in mining areas damages the surrounding environment and has various severe geological impacts. Hence, long-term monitoring of mining areas is crucial to reducing these impacts. Differential interferometric synthetic aperture radar (D-InSAR) is widely applied to monitor the subsidence in mining areas, but it cannot obtain accurate large-gradient subsidence result in the centre of the subsidence basin in mining areas due to the de-coherence phenomenon. The probability integral method (PIM) is a prediction method that can cooperate with D-InSAR (D-InSAR PIM, DPIM) to solve the problem. However, with this method, there is early convergence of the edge subsidence in narrow and long mining areas. Moreover, the PIM can only predict the spatial domain; it cannot achieve dynamic prediction. To address the above problems in the traditional DPIM data processing process, in this study, firstly, the traditional PIM was improved by adjusting the radius of the parameter and constructed an improved DPIM (IDPIM) method. The hybrid algorithm was applied to solve the parameters of the IDPIM method and then acquire subsidence results, thus solving the early convergence of edge subsidence problem characteristic of traditional PIM prediction in mining. Additionally, an area-weighting based fusion method was proposed to integrate the IDPIM results and the D-InSAR results (Area-weighting based fusion of the IDPIM and D-InSAR results, AFID) achieving whole-basin refined subsidence in mining areas. Secondly, based on a summary of subsidence laws in mining areas, the Hossfeld model was introduced and combined with the IDPIM method (IDPIM Hossfeld, IDPIM-H) to construct a subsidence dynamic prediction method. This achieved dynamic prediction of the subsidence in mining areas. A coal mine in Ordos was used as the study area, and the feasibility of the IDPIM method, the AFID method and the DPIM-H method was verified through a comparative analysis of leveling data. The results demonstrated that: (1) The results of the IDPIM method showed 8% and 66% improvement in RMSE along the striking and dip lines, respectively, over the D-InSAR results, improving the early convergence of the DPIM along the dip direction of the mine. (2) The results of the AFID method provide a 69% improvement in whole-basin RMSE over the D-InSAR results, which improves the lack of monitoring capacity in the D-InSAR technology center. (3) The results of the DPIM-H method provide a 35% improvement in basin-wide RMSE over the D-InSAR results, solving the problem of low temporal resolution of the D-InSAR technology and realizing the dynamic prediction with high temporal resolution. These findings provide a theoretical basis for future refined exploration of dynamic subsidence in mining areas.

Machine learning-enhanced modeling and characterization for optimizing dietary Fiber production from Highland barley bran.

This study investigated the modification of highland barley bran through co-fermentation of Lactobacillus bulgaricus and Kluyveromyces marxianus, and developed a dynamic prediction model for DF content under these co-fermentation conditions using machine learning algorithms. The results showed that the XGBoost algorithm could predict changes in the DF component content (R2 = 0.9553(SDF/IDF), RMSE = 0.0464.) and identify optimal fermentation conditions. Under the optimal conditions, both strains exhibited synergistic effects, where the lactic acid produced by Lactobacillus bulgaricus and β-glucosidase produced by Kluyveromyces marxianus might facilitate IDF decomposition and conversion, resulting in a maximum SDF/IDF ratio of 0.6911. This led to a 27.65 % reduction in IDF content and a 19.11 % increase in SDF content. Moreover, the physicochemical and functional properties of DF were enhanced after co-fermentation. The structure of DF became more loose and porous, its thermal stability improved, and its water-holding, oil-holding, and swelling capacities increased by 53.54 %, 16.11 %, and 44.96 %, respectively, compared with the unfermented counterpart; In terms of adsorption characteristics, its glucose, cholesterol and nitrite adsorption capacities were also significantly improved. According to in vitro gastrointestinal simulated digestion, digestion would have a great impact on the fermented DF, which showed good antioxidant properties during the intestinal digestion stage.

Prognostic Factors of Long-Term Survival and Conditional Survival Analysis in MPM Patients Treated with CRS+HIPEC: A Retrospective Study of Two Centers.

This study was designed to explore the survival benefit factors of malignant peritoneal mesothelioma (MPM) patients after cytoreductive surgery (CRS) plus hyperthermic intraperitoneal chemotherapy (HIPEC) and to make dynamic survival prediction by conditional survival (CS). Data of 212 patients with MPM who underwent CRS+HIPEC were retrospectively analyzed. Patients were divided into long-term survival (LTS) group (≥48.0 months) and short-term survival (STS) group (≤16.0 months) according to OS. Conditional survival is the probability of surviving y years after already survived for x years. Univariate and multivariate analyses were performed to explore the favorable factors of LTS. Conditional survival and Kaplan-Meier were applied to assess the postoperative survival probability. Ninety patients were enrolled: 53 (58.9%) were LTS, and 37 (41.1%) were STS. Univariate analysis revealed 14 meaningful factors (P < 0.05): age, surgery history, Karnofsky performance status, pathological types, tumor vascular emboli, lymphatic metastasis, Ki-67 index, preoperative CA125 level, peritoneal cancer index (PCI), completeness of cytoreduction, bleeding, red blood cell (RBC) transfusion, ascites, and severe adverse events (SAEs). Multivariate analysis identified that PCI ≤ 20, less RBC transfusion and no SAEs were independent prognostic factors for LTS. Five-year CS increased from 27% at 0 years to 84% at 4 years with the increasing number of survival years. The survival curve flattens at postoperative 5 years. The key factors in CRS+HIPEC for MPM patients to achieve LTS are lower tumor burden, less intraoperative RBC transfusion, and prevention of SAEs. Malignant peritoneal mesothelioma patients demonstrated a substantial increase in CS over time. Some patients may achieve clinical cure 5 years after surgery.

Machine learning for early dynamic prediction of functional outcome after stroke.

Prediction of outcome after stroke is critical for treatment planning and resource allocation but is complicated by fluctuations during the first days after onset. We propose a machine learning model that can provide hourly predictions based on the integration of continuous variables acquired within 72 h of hospital admission. We analyzed 2492 admissions for ischemic stroke in the Geneva University Hospital from 01.01.2018 to 31.12.2021, amounting to 2'131'752 unique data points. We developed a transformer model that continuously included clinical, physiological, imaging, and biological data recorded within 72 h of admission. This model was trained to generate hourly predictions of mortality and morbidity. Shapley additive explanations were used to identify the most relevant predictors to explain outcomes for each patient. The MIMIC-III database was used for external validation. Our transformer model predicts mortality, with an area under the receiver operating characteristic curve of 0.830 (95% CI 0.763-0.885) on admission, reaching 0.893 (95% CI 0.839-0.933) 72 h later for a 3-month outcome. Validated in an independent cohort, it outperforms all static models. Based on their mean explanatory weights, the top predictors included continuous clinical evaluation, baseline patient characteristics, timing from admission to acute treatment, and markers of inflammation and organ dysfunction. The performance of our transformer model demonstrates the potential of machine learning models integrating clinical, physiological, imaging, and biological variables over time after stroke. The clinical applicability of our model is further strengthened by access to hourly updated predictions along with accompanying explanations.

Real-time ocean wave prediction in time domain with autoregression and echo state networks

This study evaluates the potential of applying echo state networks (ESN) and autoregression (AR) for dynamic time series prediction of free surface elevation for use in wave energy converters (WECs). The performance of these models is evaluated on time series data at different water depths and wave conditions, including both measured and simulated data with a focus on real-time prediction of ocean waves at a given location without resolving for the surrounding ocean surface, in other words, short-time single-point forecasting. The work presented includes training the models on historical wave data and testing their ability to predict phase-resolved future surface wave patterns for short-time forecasts. Additionally, this study discusses the feasibility of deploying these models for extended time intervals. It provides valuable insights into the trade-offs between accuracy and practicality in the real-time implementation of predictive models for wave elevation, which are needed in wave energy converters to optimise the control algorithm.

Casting a sustainable future: a study on dynamic prediction and influencing factors of economic resilience in fisheries management

Marine fisheries are a critical component of the marine economy. Examining the changes in the economic resilience of marine fisheries allows for identifying potential development risks and supports informed decision-making to optimize growth patterns and mitigate risks. This study applied the CRITIC-based combined weights model to calculate China’s Marine Fisheries Economic Resilience (MFER). It investigated the temporal and spatial evolution of MFER and analyzed the influencing factors using the Geographical Detector and a Two-way fixed-effects model. Furthermore, the future development trends of MFER were predicted using the Markov chain. The results show that the overall MFER in China’s coastal provinces increased steadily from 2001 to 2020. Second, MFER exhibits a clear pattern, with low-value regions improving and high-value areas shifting toward medium- and low-value provinces. Fourth, the trajectory of the MFER center is moving from north to south, with high-value regions expanding in the same direction. Fourth, factors such as EDL, TEC, SPE, and SYS positively impact MFER, while OPE has a negative effect. Finally, the results of the Markov chain indicate that the current development state of MFER is relatively stable, with evidence of club convergence. However, the probability of maintaining the current development level is gradually decreasing. Based on these findings, provide suggestions for the preservation of marine economic security.

Simulating Ips typographus L. outbreak dynamics and their influence on carbon balance estimates with ORCHIDEE r8627

Abstract. New (a)biotic conditions resulting from climate change are expected to change disturbance dynamics, such as windthrow, forest fires, droughts, and insect outbreaks, and their interactions. These unprecedented natural disturbance dynamics might alter the capability of forest ecosystems to buffer atmospheric CO2 increases, potentially leading forests to transform from sinks into sources of CO2. This study aims to enhance the ORCHIDEE land surface model to study the impacts of climate change on the dynamics of the bark beetle, Ips typographus, and subsequent effects on forest functioning. The Ips typographus outbreak model is inspired by previous work from Temperli et al. (2013) for the LandClim landscape model. The new implementation of this model in ORCHIDEE r8627 accounts for key differences between ORCHIDEE and LandClim: (1) the coarser spatial resolution of ORCHIDEE; (2) the higher temporal resolution of ORCHIDEE; and (3) the pre-existing process representation of windthrow, drought, and forest structure in ORCHIDEE. Simulation experiments demonstrated the capability of ORCHIDEE to simulate a variety of post-disturbance forest dynamics observed in empirical studies. Through an array of simulation experiments across various climatic conditions and windthrow intensities, the model was tested for its sensitivity to climate, initial disturbance, and selected parameter values. The results of these tests indicated that with a single set of parameters, ORCHIDEE outputs spanned the range of observed dynamics. Additional tests highlighted the substantial impact of incorporating Ips typographus outbreaks on carbon dynamics. Notably, the study revealed that modeling abrupt mortality events as opposed to a continuous mortality framework provides new insights into the short-term carbon sequestration potential of forests under disturbance regimes by showing that the continuous mortality framework tends to overestimate the carbon sink capacity of forests in the 20- to 50-year range in ecosystems under high disturbance pressure compared to scenarios with abrupt mortality events. This model enhancement underscores the critical need to include disturbance dynamics in land surface models to refine predictions of forest carbon dynamics in a changing climate.

Assessment and Dynamic Prediction of Green Space Ecological Service Value in Guangzhou City, China

As an important part of the urban ecosystem, urban green space provides a variety of ecosystem services, including climate regulation, soil conservation, carbon sink and oxygen release, and biodiversity protection. However, existing remote sensing evaluation methods for ecological service value lack the evaluation indicators of ecosystem service value for Guangzhou, China, and the evaluation method depends on the land cover type. Based on remote sensing technology and random forest algorithm, this study addresses these gaps by integrating remote sensing technology with a random forest algorithm to enhance the accuracy and rationality of ESV assessments. Focusing on Guangzhou, China, we improved the ecological service value evaluation system and conducted dynamic predictions based on land-use change scenarios. Our results indicate that the total ESV of Guangzhou’s green space was USD 7.323 billion in 2020, with a projected decline to USD 6.496 billion by 2030, representing a 12.37% reduction due to urbanization-driven land-use changes. This research highlights the noticeable role of green spaces in urban sustainability and provides robust, data-driven insights for policymakers to design more effective green space protection and management strategies. The improved assessment framework offers a novel approach for accurately quantifying urban ecosystem services and predicting future trends.

Advanced obstacle recognition and trajectory prediction for autonomous vehicles using 3D CNN and GRU

Abstract. Recent years have seen a rapid advancement in new energy vehicle technology, which has increased consumer demand for autonomous driving systems. However, current self-driving cars still face many technical barriers. Obstacle detection and obstacle trajectory prediction in dynamic environments is one of the main difficulties of the technique. To improve self-driving cars' capacity for route planning and decision-making in dynamic contexts, this thesis proposes an obstacle identification and trajectory prediction model that combines a gated recurrent unit (GRU) network with a three-dimensional convolutional neural network (3D CNN). The integrated framework utilises the feature recognition capability of 3D CNN and the prediction capability of GRU to accurately predict dynamic obstacles, thereby improving the performance of the model in various types of environments (e.g., visual SLAM). Experimental results based on publicly available datasets (e.g., KITTI and Argoverse) show that the proposed method has significant advantages in dynamic obstacle identification and prediction as well as overall model operation efficiency. The findings show that the integrated framework is useful and successful in advancing the creation of self-driving automobiles.

Early Viral Dynamics Predict HIV Post-Treatment Control After Analytic Treatment Interruption.

A key research priority for developing an HIV cure strategy is to define the viral dynamics and biomarkers associated with sustained post-treatment control. The ability to predict the likelihood of sustained post-treatment control or non-control could minimize the time off antiretroviral therapy (ART) for those destined to not control and anticipate longer periods off ART for those destined to control. Mathematical modeling and machine learning were used to characterize virologic predictors of long-term virologic control using viral kinetics data from several studies in which participants interrupted ART. Predictors of post-ART outcomes were characterized using data accumulated from the time of treatment interruption, replicating real-time data collection in a clinical study, and classifying outcomes as either post-treatment control (plasma viremia ≤400 copies/mL at 2 of 3 time points for ≥24 weeks) or non-control. Potential predictors of virologic control were the time to rebound, the rate of initial rebound, and the peak plasma viremia. We found that people destined to be non-controllers could be identified within 3 weeks of rebound (prediction scores: accuracy, 80%; sensitivity, 82%; specificity, 71%). Given the widespread use of analytic treatment interruption in cure-related trials, these predictors may be useful to increase the safety of analytic treatment interruption through the early identification of people who are unlikely to become post-treatment controllers.

Characterization and prediction of hydraulic properties of traffic-compacted forest soils based on soil information and traffic treatments

Key messageThe hydraulic properties of compacted and rutted soils were evaluated through in-situ infiltration experiments and predicted based on soil texture class and traffic treatments. A significant decrease in saturated soil water content and soil hydraulic conductivity at saturation was observed. The resulting soil hydraulic parameters, when integrated into a soil water transfer model, effectively simulated water dynamics in these impacted forest soils, providing a crucial first step toward developing decision support tools for real-time trafficability. This approach can assist forest managers in minimizing the extent of soil compaction.ContextTo overcome trafficability issues of forest soils induced by heavy logging machinery, planning support tools are needed to determine suitable soil moisture conditions for traffic.AimsThis study aimed to identify the soil properties that differ significantly between undisturbed and compacted soils and to provide several estimation tools to predict the hydraulic properties of compacted soils beneath the skid trails.MethodsFour hundred seventeen water infiltration tests were conducted on 19 forest sites, mostly in North-eastern France, and analysed with the BEST method to estimate the hydraulic properties of the skid trails and undisturbed soils. The hydraulic properties of the skid trails were predicted thanks to linear mixed effect models using a bulk treatment effect, a site effect, or a skid trail degradation score. The predicted hydraulic properties were tested using a water flow model to assess their relevance regarding the prediction of water dynamics in skid trails.ResultsThe compaction effect was only significant for the logarithm of the hydraulic conductivity at saturation (log10(Ksat)) and the soil water content at saturation (θsat). For the skid trails, θsat was reduced by - 0.02 and − 0.11 m3m−3 in the 0 − 10 cm and 15 − 25 cm layers respectively, compared to undisturbed topsoil (0 − 10 cm). log10(Ksat) was reduced by − 0.38 and − 0.85 for skid trails in the 0 − 10 and 15 − 25 cm soil layers respectively, compared to undisturbed topsoil. The use of a pedotransfer function, in replacement of water infiltration tests, and their combination with the same correction coefficients proved to efficiently simulate the difference in water dynamics between skid trails and undisturbed forest soils.ConclusionEstimation of soil hydraulic properties based on in situ water infiltration experiments proved efficient to simulate water dynamics in compacted and rutted forest soils. Yet, further studies are needed to identify the most adapted pedotransfer function to forest soils and to test the generalisation of our findings in different conditions, especially deeply rutted soils (rut depths > 12 cm).

Dynamic Adaptability and Prediction Optimization in Power Material Supply Chains Using XGBoost

Dynamic Adaptability and Prediction Optimization in Power Material Supply Chains Using XGBoost

Multi-task Bayesian model combining FDG-PET/CT imaging and clinical data for interpretable high-grade prostate cancer prognosis.

We propose a fully automatic multi-task Bayesian model, named Bayesian Sequential Network(BSN), for predicting high-grade(Gleason 8) prostate cancer(PCa) prognosis using pre-prostatectomy FDG-PET/CT images and clinical data. BSN performs one classification task and five survival tasks: predicting lymph node invasion (LNI), biochemical recurrence-free survival (BCR-FS), metastasis-free survival, definitive androgen deprivation therapy-free survival, castration-resistant PCa-free survival, and PCa-specific survival(PCSS). Experiments are conducted using a dataset of 295patients. BSN outperforms widely used nomograms on all tasks except PCSS, leveraging multi-task learning and imaging data. BSN also provides automated prostate segmentation, uncertainty quantification, personalized feature-based explanations, and introduces dynamic predictions, a novel approach that relies on short-term outcomes to refine long-term prognosis. Overall, BSN shows great promise in its ability to exploit imaging and clinicopathological data to predict poor outcome patients that need treatment intensification with loco-regional or systemic adjuvant therapy for high-risk PCa.

AI-integrated network for RNA complex structure and dynamic prediction.

RNA complexes are essential components in many cellular processes. The functions of these complexes are linked to their tertiary structures, which are shaped by detailed interface information, such as binding sites, interface contact, and dynamic conformational changes. Network-based approaches have been widely used to analyze RNA complex structures. With their roots in the graph theory, these methods have a long history of providing insight into the static and dynamic properties of RNA molecules. These approaches have been effective in identifying functional binding sites and analyzing the dynamic behavior of RNA complexes. Recently, the advent of artificial intelligence (AI) has brought transformative changes to the field. These technologies have been increasingly applied to studying RNA complex structures, providing new avenues for understanding the complex interactions within RNA complexes. By integrating AI with traditional network analysis methods, researchers can build more accurate models of RNA complex structures, predict their dynamic behaviors, and even design RNA-based inhibitors. In this review, we introduce the integration of network-based methodologies with AI techniques to enhance the understanding of RNA complex structures. We examine how these advanced computational tools can be used to model and analyze the detailed interface information and dynamic behaviors of RNA molecules. Additionally, we explore the potential future directions of how AI-integrated networks can aid in the modeling and analyzing RNA complex structures.

Reducing uncertainty with iterative model updating parses effects of competition and environment on salamander occupancy.

Making timely management decisions is often hindered by uncertainty. Monitoring reduces two key types of uncertainty. First, it serves to reduce structural uncertainty of how the system works and provides support for expectations of how a system works. Second, it serves to reduce parametric uncertainty of the drivers of system dynamics. By combining monitoring data and quantitative models, we can reduce structural and parametric uncertainty. To demonstrate this, we focus on the Shenandoah salamander (Plethodon shenandoah), a United States Federally Endangered Species. Early work suggested that P. shenandoah extinction risk results from competition with a conspecific (Plethodon cinereus). However, more recent work has found equivocal support for this claim, instead suggesting that abiotic factors, such as moisture and temperature, drive P. shenandoah persistence. Using long-term monitoring data, we find that while competition may play a part in P. shenandoah extinction risk, measures of surface moisture are better predictors of occupancy dynamics. Further, we find decreased detection rates of P. shenandoah when P. cinereus is present, suggesting a conflation of detection probability with actual competition, which cautions against making inference from unadjusted observations of occurrence. Using multiple lines of inquiry allows for more robust understanding of system drivers in the face of high uncertainty, increasing opportunities to manage extinction risk.

Spike and Slab Regression for Nonstationary Gaussian Linear Mixed Effects Modeling of Rapid Disease Progression

ABSTRACTSelect measures of social and environmental determinants of health (referred to as “geomarkers”), predict rapid lung function decline in cystic fibrosis (CF), defined as a prolonged decline relative to patient and/or center‐level norms. The extent to which hyper‐localization, defined as increasing the spatiotemporal precision of geomarkers, aids in prediction of rapid lung decline remains unclear. Linear mixed effects (LME) models with specialized covariance functions have been used for predicting rapid lung function decline, but there are few options to properly incorporate spatial correlation into the covariance functions while inducing simultaneous variable selection. Our innovative Bayesian model uses a spike and slab prior for simultaneous variable selection and offers additional advantages when coupled with nonstationary Gaussian LME modeling. This model also incorporates spatial correlation through an additional random effect term that accounts for spatial correlation based on ZIP code distances. We validated the model with simulations and applied it to real CF data from a Midwestern CF Center. We demonstrate how a combination of demographic, clinical, and geomarker variables can be selected as optimal predictors using Bayesian false discovery rate controlling rule. Our results indicate that incorporating spatiotemporal effects and geomarkers into this novel Bayesian stochastic LME model enhances the dynamic prediction of rapid CF disease progression.

Liver stiffness as a dynamic predictor of decompensation and mortality in alcohol-related liver disease

Liver stiffness as a dynamic predictor of decompensation and mortality in alcohol-related liver disease