Statistical Prediction Research Articles

Second Moment Closure Modeling and Direct Numerical Simulation of Stratified Shear Layers

AbstractBuoyant shear layers encountered in many engineering and environmental applications have been studied by researchers for decades. Often, these flows have high Reynolds and Richardson numbers, which leads to significant/intractable space–time resolution requirements for direct numerical simulation (DNS) or large eddy simulation (LES). On the other hand, many of the important physical mechanisms, such as stress anisotropy, wake stabilization, and regime transition, inherently render eddy viscosity-based Reynolds-averaged Navier–Stokes (RANS) modeling inappropriate. Accordingly, we pursue second-moment closure (SMC), i.e., full Reynolds stress/flux/variance modeling, for moderate Reynolds number nonstratified, and stratified shear layers for which DNS is possible. A range of submodel complexity is pursued for the diffusion of stresses, density fluxes and variance, pressure strain and scrambling, and dissipation. These submodels are evaluated in terms of how well they are represented by DNS in comparison to the exact Reynolds-averaged terms, and how well they impact the accuracy of full RANS closure. For the nonstratified case, SMC model predicts the shear layer growth rate and Reynolds shear stress profiles accurately. Stress anisotropy and budgets are captured only qualitatively. Comparing DNS of exact and modeled terms, inconsistencies in model performance and assumptions are observed, including inaccurate prediction of individual statistics, non-negligible pressure diffusion, and dissipation anisotropy. For the stratified case, shear layer and gradient Richardson number growth rates, and stress, flux and variance decay rates, are captured with less accuracy than corresponding flow parameters in the nonstratified case. These studies lead to several recommendations for model improvement.

Machine learning and statistical prediction of fastball velocity with biomechanical predictors

In recent years, one of the most important factors for success among baseball pitchers is fastball velocity. The purpose of this study was to (1) to develop statistical and machine learning models of fastball velocity, (2) to identify the strongest predictors of fastball velocity, and (3) to compare the models' prediction performances. Three dimensional biomechanical analyses were performed on high school (n = 165) and college (n = 62) baseball pitchers. A total of 16 kinetic and kinematic predictors from the entire pitching sequence were included in regression and machine learning models. All models were internally validated through ten-fold cross-validation. Model performance was evaluated through root mean square error (RMSE) and calibration with 95% confidence intervals. Gradient boosting machines demonstrated the best prediction performance [RMSE: 0.34; Calibration: 1.00 (95% CI: 0.999, 1.001)], while regression demonstrated the greatest prediction error [RMSE: 2.49; Calibration: 1.00 (95% CI: 0.85, 1.15)]. Maximum elbow extension velocity (relative influence: 19.3%), maximum humeral rotation velocity (9.6%), maximum lead leg ground reaction force resultant (9.1%), trunk forward flexion at release (7.9%), time difference of maximum pelvis rotation velocity and maximum trunk rotation velocity (7.8%) demonstrated the greatest influence on pitch velocity. Gradient boosting machines demonstrated better calibration and reduced RMSE compared to regression. The influence of lead leg ground reaction force resultant and trunk and arm kinematics on pitch velocity demonstrates the interdependent relationship of the entire kinetic chain during the pitching motion. Coaches, players, and performance professionals should focus on the identified metrics when designing pitch velocity improvement programs.

The influence of laser directed energy deposition (DED) processing parameters for Al5083 studied by central composite design

Laser directed energy deposition (DED) of high-quality structural Al-based alloys is challenging due to the inherent physical and thermal properties of the Al powder feedstock. Therefore, an in-depth understanding of the influence of the applied processing parameters on the characteristics of the deposited material is paramount if one is to attain optimal performance. The objective of this study is to investigate the influence of the dominant processing parameters (laser power, scan speed, powder mass flow rate (PMFR), and hatch spacing) on the geometrical characteristics (track's height and dilution) of Al5083 double tracks fabricated using Laser Engineered Net Shaping (LENS®). Central composite design (CCD) response surface methodology (RSM) was utilized to study the influence of the varied processing parameters and their interactions and to develop an empirical statistical prediction model for the studied responses. The results reveal that the applied PMFR has a strong influence on the deposited track's height (positive) and dilution (negative). The laser power at the first-order factor shows a weak influence for both studied responses. However, the interaction between the laser power and the hatch spacing shows significant effects on the track's height and dilution. The microstructure and microhardness of the as-deposited Al5083 double tracks are also discussed, as proxies to the anticipated performance of the deposited material.

Statistical creep failure time of unidirectional carbon fiber reinforced plastic under bending load

Our developed accelerated testing methodology (ATM) based on the matrix resin viscoelasticity for the creep and fatigue failure life prediction of fiber reinforced polymers (FRP) was applied to the statistical prediction of long-term creep failure life for the longitudinal bending of unidirectional Carbon fiber reinforced plastic (CFRP) laminates which is an important basic item for the durability design of CFRP structures used for aircraft and others. As results, the statistical creep failure times measured under several constant bending loads at an arbitrary temperature for unidirectional CFRP laminates were agreed with the predicted results obtained by substituting the matrix resin viscoelasticity and the flexural static strengths of CFRP laminates statistically and easily measured at various temperatures into the formulation of ATM. The long-term creep strength under bending load at an arbitrary temperature predicted were compared with that under tension load obtained by our previous paper. It was clear that the creep strength under bending load degreases drastically with increase in time and temperature comparing with that under tension load; therefore, the effect of time and temperature on the creep failure life under bending load is larger than that under tension load.

Development and evaluation of the bootstrap resampling technique based statistical prediction model for Covid-19 real time data : A data driven approach

The objective of the article is to develop ‘earlyR’ package based novel coronavirus disease (COVID-19) forecasting model. The reported COVID-19 serial interval data is applied for obtaining maximum likelihood value of the reproduction number (R0) using maximum likelihood approach and ‘projections’ package is applied for getting trajectories of epidemic curve. The minimum, median, mean and maximum projected value of R0 with 95% confidence interval (CI) is obtained by using bootstrap resampling strategy and the predicted cumulative probable count of new cases is also presented with different quantile. To validate the results with real scenario, the past COVID-19 data is considered. The % error rate ranges from -7.91% to 21.27% for the developed model for the five Indian States.

Spatial–temporal distribution and forecasting model of precipitation using dynamic-statistical information fusion

Abstract In order to explore the evolvement mechanism of hydrometeorological elements, spatial–temporal distribution of precipitation in the Huai river basin is studied by statistical drawing and empirical orthogonal function decomposition. How to make an objective combination for the predictive results of precipitation? Information fusion in data assimilation is introduced to merge the improved National Centers for Environmental Prediction coupled forecast system model version 2 (CFSv2) with the multilinear regression model. Firstly, in terms of time, the annual precipitation is apt to decline at most stations within 30 years, and precipitation mainly concentrates in the flood season. The characteristics of spatial distribution are similar to topographic features. It can also be found that precipitation gradually decreases from south to north. Secondly, from statistical forecasting, the relationship between precipitation and global sea surface temperature (SST) is explored. Prediction equation is established with SST and the average precipitation. Thirdly, from dynamic model forecasting, the CFSv2 original model and the CFSv2 statistical downscaling model are used to analyze the influence of model deviation on fusion prediction. The optimum interpolation assimilation method is applied for realizing the optimal integration of statistical and dynamic model prediction. Finally, the standardized precipitation index (SPI) is calculated by the combined forecasting of annual precipitation to evaluate drought conditions. The results show that SST is an important factor affecting precipitation, which may be applied as a forecasting direction with other factors. The merged precipitation prediction skill by the CFSv2 original model and the statistical model do not have the great promotion, which is still lower than the prediction skill only by the statistical model. However, the merged precipitation prediction skill by the CFSv2 statistical downscaling model and the statistical model is better than the prediction skills by the two models mentioned above, respectively. These indicate that when the prediction difference between the models is large, the merged prediction error cannot be minimized. When the prediction skill levels are equal, there is an improvement in the merged result. So, it is necessary to revise the climate dynamic model by downscaling. What is more, the obtained drought levels match the actual disaster conditions, providing theoretical support of hydrology and meteorology for the prevention of natural disasters.

Effectiveness of a novel polyaniline@Fe-ZSM-5 hybrid composite for Orange G dye removal from aqueous media: Experimental study and advanced statistical physics insights

A polyaniline@Fe-ZSM-5 composite was synthesized via an in situ interfacial polymerization procedure. The morphology, crystallinity, and structural features of the as-developed PANI@Fe-ZSM-5 composite were assessed using scanning electron microscopy – energy dispersive spectroscopy (SEM-EDS), X-ray diffraction (XRD), Fourier-transform infrared (FTIR) spectroscopy, and X-ray photoelectron spectroscopy (XPS). The composite was efficiently employed for the first time as an adsorbent Orange G (OG) dyestuff from water. The OG dye adsorption performance was investigated as a function of several operating conditions. The kinetic study demonstrated that a pseudo-second-order model was appropriate to anticipate the OG adsorption process. The maximum adsorption capacity was found to be 217 mg/g. The adsorption equilibrium data at different temperatures were calculated via advanced statistical physics formalism. The entropy function indicated that the disorder of OG molecules improved at low concentrations and lessened at high concentrations. The free enthalpy and internal energy functions suggested that the OG adsorption was a spontaneous process and physisorption in nature. Regeneration investigation showed that the PANI@Fe-ZSM-5 could be effectively reused up to five cycles. The main results of this work provided a deep insight on the experimental study supported by advanced statistical physics prediction for the adsorption of Orange G dye onto the novel polyaniline@Fe-ZSM-5 hybrid composite. Additionally, the experimental and advanced statistical physics findings stated in this study may arouse research interest in the field of wastewater treatment.

A Light-Weight Statistical Latency Measurement Platform at Scale

The statistical value of latencies between two sets of hosts over a given period, which is referred as to the statistical latency, can benefit many applications in the next-generation networks, for example, Network-in-a-Box-based resource provisioning. However, the existing methods can hardly achieve low measurement cost and high prediction accuracy simultaneously in large-scale scenarios. In this article, we design a light-weight statistical latency measurement platform named DMS (DNS-based statistical latency Measurement platform at Scale). DMS achieves high measurement accuracy by introducing a metric space to select the closest open recursive DNS (Domain Name System) server to a given host, and predicting the end-to-end latency between two hosts via the measured latency between the two corresponding DNS servers. To reduce the overall measurement overhead, DMS clusters the hosts in the metric space with the open recursive DNS infrastructure in the network as the cluster center, thus achieving low measurement cost and good scalability in large scale simultaneously. To evaluate the performance of DMS, we implement a prototype system in the network. Compared to the widely adopted method King, DMS can reduce the relative error by 18.5% for real-time end-to-end latency prediction and 33% for statistical latency prediction.

MODELLING THE AIR QUALITY INDEX FOR BOLU, TURKEY

The monthly air quality index (AQI) derived from ground observation stations that obtained daily air pollutants information for 1990- through 2010 was analyzed in this study. AQI was evaluated using the common comparative index method presented by the U.S. Environmental Protection Agency (USEPA), and a statistically based approach was used for predicting the AQI value. With the first method, AQI was predicted using the USEPA subindex formula for different pollutants, such as particulate matter and sulfur dioxide, which contribute the most to air pollution. A combination of the principal component analysis (PCA) and multiple linear regression (MLR) methods were used with the measured values of climate variables obtained from the ground stations for the most effective contributors and a prediction was modelled. The results of these two methods were compared and evaluated for consistency. Two methods were presented for determining the AQI value. According to the findings, the common comparative index method was consistent with the statistical prediction models, and the best results were obtained using PCA models with varimax rotation.

Offshore renewable energy site correlated wind-wave statistics

Offshore engineering often requires estimation of extreme wind speeds and wave heights at offshore in-situ locations. This paper presents practical approach to study extreme wind speed and wave height statistics, based on available in-situ hourly wind speed and wave height maxima. The wind and wave data, studied in this paper, was obtained from numerical hind cast model, locally applied to the offshore the area that covers SEM-REV (European wind and wave energy research site) location, for the time period 2001–2010 years.To obtain accurate local wind and wave data set, accurate wave and wind measurements from in-situ monitoring tools (meteorological stations and buoys) as well as remote satellite sensing data (ENVISAT and TOPEX satellite records) were plugged into atmospheric wind and wave model.The novel bivariate correction technique based on the Average Conditional Exceedance Rate (ACER) method has been presented in brief detail. The bivariate correction method produced quite accurate extreme value predictions, efficiently utilizing available wind speeds and wave heights data set.In some practical situations it would be useful to improve accuracy of some statistical predictions, using information supplied by another synchronous highly correlated random process that has been measured for a longer time than the process of interest. In this paper the novel technique of improving correlated extreme wind speed and wave height predictions has been presented.

A Statistical Prediction Model of River Flood

A Statistical Prediction Model of River Flood

Parametric analysis and new performance correlation of the surface conventional rigid pavement temperature

A statistical model for predicting pavement surface temperature has been developed for the first time in this study. The main objective of this work is to analyse the impact of the input factors (solar radiation, ambient temperature, wind speed, and albedo) most affecting the conventional rigid pavement surface temperature. A multi-physics numerical model combining radiative, conductive, and convective heat transfers with a response surface methodology (RSM) model was created as an innovative and integrated approach. The numerical model of two-dimensional heat transport based on volume finite techniques is performed using FORTRAN software. Moreover, the analysis of variance (ANOVA) is used to establish the polynomial model or new correlation (quadratic regression) and analyse each input factor's statistical importance. Heat transfer model findings were in good agreement with the experimental data. For pavement surface temperature, the results show that the adjusted Radj2and predicted determination coefficients R2 achieved are 0.9912 and 0.9604, respectively. A good agreement is obtained between the statistical model's prediction values and the heat transfer model's numerical data. Based on the four analysed input factors, a simple polynomial statistical model is developed to forecast and reduce the conventional rigid surface pavement temperature. A wind speed of 1.215 m/s, an air temperature of 293.634 K, solar radiation of 222.305 W/m2 and albedo of 0.782 were found to be the optimal operating conditions for obtaining the lowest possible temperature on the pavement's surface.

Physics-based Residual Kriging for dynamically evolving functional random fields

We present a novel approach named Physics-based Residual Kriging for the statistical prediction of spatially dependent functional data. It incorporates a physical model—expressed by a partial differential equation—within a Universal Kriging setting through a geostatistical modelization of the residuals with respect to the physical model. The approach is extended to deal with sequential problems, where samples of functional data become available along consecutive time intervals, in a context where the physical and stochastic processes generating them evolve, as time intervals succeed one another. An incremental modeling is used to account for both these dynamics and the misfit between previous predictions and actual observations. We apply Physics-based Residual Kriging to forecast production rates of wells operating in a mature reservoir according to a given drilling schedule. We evaluate the predictive errors of the method in two different case studies. The first deals with a single-phase reservoir where production is supported by fluid injection, while the second considers again a single-phase reservoir but the production is driven by rock compaction.

Bootstrapping Time-Varying Uncertainty Intervals for Extreme Daily Return Periods

This study aims to overcome the problem of dimensionality, accurate estimation, and forecasting Value-at-Risk (VaR) and Expected Shortfall (ES) uncertainty intervals in high frequency data. A Bayesian bootstrapping and backtest density forecasts, which are based on a weighted threshold and quantile of a continuously ranked probability score, are developed. Developed backtesting procedures revealed that an estimated Seasonal autoregressive integrated moving average-generalized autoregressive score-generalized extreme value distribution (SARIMA–GAS–GEVD) with a skewed student-t distribution had the best prediction performance in forecasting and bootstrapping VaR and ES. Extension of this non-stationary distribution in literature is quite complicated since it requires specifications not only on how the usual Bayesian parameters change over time but also those with bulk distribution components. This implies that the combination of a stochastic econometric model with extreme value theory (EVT) procedures provides a robust basis necessary for the statistical backtesting and bootstrapping density predictions for VaR and ES.

A deep generative approach for crash frequency model with heterogeneous imbalanced data

Crash frequency model is often subject to excessive zero observation because of the rare nature of crashes. To address the problem of imbalanced crash data, a deep generative approach – augmented variational autoencoder – was proposed to generate synthetic crash data for the association measure between crash and possible explanatory factors. This approach was characterized by a factorized generative model and refined objective function. For instance, the generative model can handle heterogeneous data including real-valued, nominal and ordinal distributions. On the other hand, the refined objective function can control for the random effect by better recognizing both the zero-crash and non-zero crash cases. In this study, comprehensive traffic and crash data of multiple distribution types in Hong Kong in the period between 2014 and 2016 were used. To assess the data generation performance of the proposed augmented variational autoencoder method, a conventional data synthesis technique (synthetic minority oversampling technique-nominal continuous) was also considered. Performances of crash frequency models of total crashes and fatal and severe injury crashes are assessed. For total crashes, the results of parameter estimation, in terms of statistical fit, prediction accuracy, and explanatory factors identified, of the crash frequency model based on synthetic data using the augmented variational autoencoder method adhered closer to that based on original data, compared to that based on synthetic data using the synthetic minority oversampling technique-nominal continuous method. For fatal and severe injury crashes, zero-crash observations were prevalent, with the ratio of zero-crash to non-zero crash cases of 9 to 1. Crash data was first balanced using the proposed augmented variational autoencoder method. Then, fatal and severe injury crash frequency models using correlated random parameter models based on original data and balanced data were estimated respectively. Results indicate that fatal and severe injury crash frequency model based on balanced data outperforms its counterpart, with the lowest root mean square error, lowest mean absolute error, and highest number of crash explanatory factors identified. More importantly, correlation between the random parameters can be revealed. Findings of this study should shed light to both researchers and practitioners for the development of crash frequency models, with which the problem of excessive zero observations is prevalent when highly disaggregated traffic and crash data by time and space are used.

Reliability as a Precondition for Trust-Segmentation Reliability Analysis of Radiomic Features Improves Survival Prediction.

Machine learning results based on radiomic analysis are often not transferrable. A potential reason for this is the variability of radiomic features due to varying human made segmentations. Therefore, the aim of this study was to provide comprehensive inter-reader reliability analysis of radiomic features in five clinical image datasets and to assess the association of inter-reader reliability and survival prediction. In this study, we analyzed 4598 tumor segmentations in both computed tomography and magnetic resonance imaging data. We used a neural network to generate 100 additional segmentation outlines for each tumor and performed a reliability analysis of radiomic features. To prove clinical utility, we predicted patient survival based on all features and on the most reliable features. Survival prediction models for both computed tomography and magnetic resonance imaging datasets demonstrated less statistical spread and superior survival prediction when based on the most reliable features. Mean concordance indices were Cmean = 0.58 [most reliable] vs. Cmean = 0.56 [all] (p < 0.001, CT) and Cmean = 0.58 vs. Cmean = 0.57 (p = 0.23, MRI). Thus, preceding reliability analyses and selection of the most reliable radiomic features improves the underlying model’s ability to predict patient survival across clinical imaging modalities and tumor entities.

Influence on kinesiophobia by disability, physical, and behavioural variables after a heart transplantation.

From clinical experience, kinesiophobia represents a barrier to being physically active after a heart transplantation (HTx), but studies in this field are lacking. Identifying the factors associated with kinesiophobia is essential to determine preventive interventions to avoid negative consequences for health. To study the influence of disability, physical, and behavioural variables on kinesiophobia in patients with an HTx. A total of 117 patients with an HTx [51 women; mean age 56 (SD 12.1) years] were recruited at an outpatient clinic. These patients were asked to fill in questionnaires measuring kinesiophobia, self-reported physical activity (PA), exercise self-efficacy, motivation for PA, and disability. A multiple regression analysis was conducted to examine the statistical prediction of kinesiophobia as a dependent variable, with the questionnaires, gender and education as independent variables. The independent variables explained 70% of the variance in kinesiophobia. The prediction model was significant (F = 32.1, P &lt; 0.001). The time from transplantation (standardised coefficient, beta; -0.17), the total exercise self-efficacy (-0.16), extrinsic motivation (-0.23), and the disability total score (0.63) were significant predictors of kinesiophobia, while the independent variables of gender, education, intrinsic motivation, and the PA total score were not significant. This study highlights that a short time from transplantation, low self-efficacy, low extrinsic motivation, and a high level of disability explained high levels of kinesiophobia in patients after an HTx. These results suggest that an increased awareness of the biopsychosocial health perspective is essential in order to maximising patient outcomes after an HTx.

Training Data Selection for Machine Learning-Enhanced Monte Carlo Simulations in Structural Dynamics

The evaluation of structural response constitutes a fundamental task in the design of ground-excited structures. In this context, the Monte Carlo simulation is a powerful tool to estimate the response statistics of nonlinear systems, which cannot be represented analytically. Unfortunately, the number of samples which is required for estimations with high confidence increases disproportionally to obtain a reliable estimation of low-probability events. As a consequence, the Monte Carlo simulation becomes a non-realizable task from a computational perspective. We show that the application of machine learning algorithms significantly lowers the computational burden of the Monte Carlo method. We use artificial neural networks to predict structural response behavior using supervised learning. However, one shortcoming of supervised learning is the inability of a sufficiently accurate prediction when extrapolating to data the neural network has not seen yet. In this paper, neural networks predict the response of structures subjected to non-stationary ground excitations. In doing so, we propose a novel selection process for the training data to provide the required samples to reliably predict rare events. We, finally, prove that the new strategy results in a significant improvement of the prediction of the response statistics in the tail end of the distribution.

Turbulent boundary layer flow over regularly and irregularly arranged truncated cone surfaces

Aiming to study the rough-wall turbulent boundary layer structure over differently arranged roughness elements, an experimental study was conducted on flows with regular and random roughness. Varying planform densities of truncated cone roughness elements in a square staggered pattern were investigated. The same planform densities were also investigated in random arrangements. Velocity statistics were measured via two-component laser Doppler velocimetry and stereoscopic particle image velocimetry. Friction velocity, thickness, roughness length and zero-plane displacement, determined from spatially averaged flow statistics, showed only minor differences between the regular and random arrangements at the same density. Recent a priori morphometric and statistical drag prediction methods were evaluated against experimentally determined roughness length. Observed differences between regular and random surface flow parameters were due to the presence of secondary flows which manifest as high-momentum pathways and low-momentum pathways in the streamwise velocity. Contrary to expectation, these secondary flows were present over the random surfaces and not discernible over the regular surfaces. Previously identified streamwise-coherent spanwise roughness heterogeneity does not seem to be present, suggesting that such roughness heterogeneity is not necessary to sustain secondary flows. Evidence suggests that the observed secondary flows were initiated at the front edge of the roughness and sustained over irregular roughness. Due to the secondary flows, local turbulent boundary layer profiles do not scale with local wall shear stress but appear to scale with local turbulent shear stress above the roughness canopy. Additionally, quadrant analysis shows distinct changes in the populations of ejection and sweep events.

Contemplated Investigation and Statistical Prediction of the Swelling Potential of Expansive Soils Using Index Properties of Holte Town, Southern Ethiopia

Expansive soils are problematic by exhibitingl shrink‐swell behaviors and high plasticity under different moisture conditions, which is communal in salient areas of Ethiopia. This article aims to conduct a contemplated investigation and statistical analysis‐based prediction for swelling potentials of soils using the index properties. This is used to define the properties of soils by investigation and to minimize the rigorous futuristic investigation by providing the prediction model of soils of the Holte town. Hence, thirty soil samples were collected from fifteen test pits with a depth of 1.5 m to 3.0 m from various locations. Investigations were conducted in the field and laboratory. Results revealed that the in situ moisture content, dry density, and bulk unit weight ranges were found, 31.21–40.93%, 1.20–1.30 g/cm3, and 1.60–1.73 g/cm3, respectively. The soil grain results presented that the study area is dominantly fine‐grained soil properties (i.e., silt and clay). Liquid limit, plastic limit, plasticity index, and liquidity index ranged from 58 to 77%, 26–33%, 28–45%, and 0.00–0.35%, respectively, which assured that the soils in this town have 87% high plastic clay (CH). In addition, the swelling potential, swelling pressure, free swell, and activity investigations revealed that the soils in this town are highly plastic with a high degree of expansiveness. This was found that the swelling potential is the main measuring property of expansive soils, which was predicted by the statistical tool SPSS using numerous index properties. After conducting numerous statistical models, the best swelling potential was found Sp = 6.876γd + 0.178PI − 10.664, which is maintained by 90% of the points found within the regression line having a mean confidence level of 99.99%. In conclusion, the obtained model was validated and was found as an optimal measuring tool of expansiveness through the swelling potential of soils specifically in Holte town.