A crash-prediction model for multilane roads

The coronavirus disease 2019 (COVID-19) pandemic was defined by the World Health Organization (WHO) as a global epidemic on March 11, 2020, as the infectious disease that threatens public health fatally. In this study, the main aim is to model the impact of various air pollution causes on mortality data due to the COVID-19 pandemic by Generalized Linear Mixed Model (GLMM) approach to make global statistical inferences about 174 WHO member countries as subjects in the six WHO regions. “Total number of deaths by these countries due to the COVID-19 pandemic” until July 27, 2022, is taken as the response variable. The explanatory variables are taken as the WHO regions, the number of deaths from air pollution causes per 100.000 population as “household air pollution from solid fuels,” “ambient particulate matter pollution,” and “ambient ozone pollution.” In this study, Poisson, geometric, and negative binomial (NB) regression models with “country” taken as fixed and random effects, as special cases of GLMM, are fitted to model the response variable in the aspect of the above-mentioned explanatory variables. In the Poisson, geometric, and NB regression models, Iteratively Reweighted Least Squares parameter estimation method with the Fisher-Scoring iterative algorithm under the log-link function as canonical link function is used. In the GLMM approach, Laplace approximation is also used in the prediction of random effects. In this study, six different Poisson, geometric, and NB regression models with fixed and random effects are established for 174 countries all over the world to make global statistical inferences for investigating the relationships between “total number of deaths” by these countries due to the COVID-19 pandemic and “air pollution causes.” As a result of this study, “NB mixed-effects regression model” as the most appropriate GLMM is used to make global statistical inferences about the impact of the various air pollution causes on the mortality data due to the COVID-19 pandemic.

This paper presents results of a safety analysis of a freeway managed-lane strategy, that is, a time-of-day managed-lane strategy that concurrently allows use of the inner left lanes by high-occupancy vehicles (HOVs) and use of right shoulders as general purpose travel lanes during peak hours. Recent 3-year crash data and corresponding annual average daily traffic (AADT) volumes and lane-type-specific AADT volumes were identified for various lane types, including the inner left lanes for HOV only use during peak hours, general purpose lanes, right shoulder lanes, and all lanes as a whole. Negative binomial regression models were used to estimate the effect of this traffic operations system and other factors relevant to crash frequency. The negative binomial regression model analyses present no evidence that the interest factors, including the managed-lane strategy during peak hours, AADT volumes, merging and diverging influence areas, weather, light conditions, and existence of pull-off areas, affect the crash frequency when aggregated across all lanes. The variable AADT volumes in the specific analysis of general purpose lanes appear to be significant and show about a 2% increase in weekday crashes for each increase of 1,000 vehicles per day in the AADT range of 50,000 to 83,000 vehicles per day. Right shoulder specific analysis shows that motorist behaviors at the merge and diverge areas during adverse light conditions are significant and shows an increase of about 38% in crashes in these areas. The managed-lane strategy does not appear to be significant to the crash frequency in the inner left lanes for HOV, general purpose lanes, or right shoulders.

Tujuan: Untuk menjelaskan gambaran faktor lingkungan fisik meliputi: suhu, kelembaban, kecepatan angin dan curah hujan, dan menganalisis secara spasial dan temporal pengaruhnya terhadap kejadian malaria di Kabupaten Banjarnegara tahun 2011-2015. Metode: Penelitian ini menggunakan desain studi ekologi dengan pendekatan spasial-temporal. Data yang digunakan adalah data time series bulan kejadian malaria dan faktor lingkungan fisik di Kabupaten Banjarnegara tahun 2011-2015. Model regresi poisson dan binomial negatif digunakan untuk menganalisis pengaruh faktor lingkungan fisik dengan kejadian malaria berdasrakan hasil pemilihan model terbaik dengan melihat nilai Akaike’s Information Criteria (AIC) dan Bayesian Information Criteria (BIC). Data faktor lingkungan fisik dilakukan interpolasi menggunakan metode IDW ( Inverse Distance Weighting ) pada software Sistem Informasi Geografis. Hasil: Rata-rata suhu udara dan kelembaban udara dari tahun 2011-2015 berturut-turut 25,79 o C dan 87,37%, sedangkan rata-rata kecepatan angin dan curah hujan berturut-turut 26,07 km/h, 276,02 mm. Analisis statistik, grafik/ time trend , dan spasial menunjukkan bahwa ada hubungan antara suhu, kelembaban, kecepatan angin dan curah hujan dengan kejadian malaria di Kabupaten Banjarnegara. Hasil analisis regresi binomial negatif menunjukkan bahwa secara bermakna suhu pada bulan yang sama (P= 0,036), suhu pada dua bulan sebelumnya (P=0,022) berkorelasi negatif berturut-turut dengan Estimate Value (-0,273, -0,319) sedangkan curah hujan pada satu bulan sebelumnya (P=0,0001) berkorelasi positif dengan Estimate value (0,002) terhadap kejadian malaria di Kabupaten Banjarnegara Tahun 2011-2015. Hasil analisis spasial menunjukkan bahwa kasus tertinggi malaria berada di Kecamatan Banjarmangu, Punggelan dan Pagedongan dengan kisaran suhu 25,64 o C-26,77 o C, kisaran kelembaban 86,07%-88,41%, kisaran kecepatan angin 18,71km/h-28,30km/h, dan kisaran curah hujan 225,25mm-318,24mm. Simpulan: Suhu pada bulan yang sama, dan suhu pada dua bulan sebelumnya memiliki pengaruh yang negatif terhadap kejadian malaria sedangkan curah hujan pada satu bulan sebelumnya memiliki pengaruh yang positif terhadap kejadian malaria di Kabupaten Banjarnegara. Maka, untuk mengatasi masalah kejadian malaria dan melaksanakan program penanggulangannya maka Pemerintah Daerah Kabupaten Banjarnegara perlu mempertimbangan faktor lingkungan fisik.

Infant mortality has generally been increasing and has become an issue that urgently needs to be addressed. As the number of infant deaths is count data, a Poisson regression model is needed to determine the causal factors. However, the assumption of equidispersion in Poisson regression is rarely satisfied. The overdispersion issue is frequently found in real data. Thus, this research employs mixed Poisson distribution modeling to overcome the overdispersion issue, namely, the inverse Gaussian Poisson regression (IGPR) model. In this study, a simple IGPR model, a modified IGPR model, and the negative binomial regression (NBR) model are compared. The results show that the modified IGPR model and the NBR model with an exposure variable outperform the benchmark, based on the global deviance and Akaike Information Criteria (AIC) value, to model the number of infant deaths in East Nusa Tenggara, Indonesia. The significant predictors that affect the number of infant mortalities are the percentage of complete basic immunization, the percentage of low birth weight (LBW), the percentage of babies under six months who receive exclusive breastfeeding, the percentage of infants who receive vitamin A, and the percentage of births assisted by health workers in the district.

The two most popular models inwell-known count regression models are Poisson and negative binomial regression models. Poisson regression is a generalized linear model form of regression analysis used to model count data and contingency tables. Poisson regression assumes the response variable Y has a Poisson distribution, and assumes the logarithm of its expected value can be modeled by a linear combination of unknown parameters. Negative binomial regression is similar to regular multiple regression except that the dependent (Y) variables an observed count that follows the negative binomial distribution. This research studies some factors affecting divorce using Poisson and negative binomial regression models. The factors are unemployment rate, inflation and Gini coefficient. The data were taken from the website of the Statistics Center for the years 2002 to 2023.Under the Poisson regression model, each factor has been reported to have an effect on the divorce rate. The two factors of inflation and unemployment had a direct effect and income inequality factors had an inverse effect on the divorce rate. But, under the negative binomial regression model, only inflation has an effect on the number of divorces. It is worth noting that according to the AIC values, the negative binomial regression model has a better fit than the Poisson regression model because its AIC value is lower.

Nepal has experienced the glacial reduction of under-five mortality in recent years which portends greater challenges to achieve the global target outlined in the Sustainable Development Goals (SDGs) for reducing childhood mortality. To resolve this inevitable adversity and safeguard newborns' lives, additional scientific studies are necessary to plan for evidence-based interventions. This study aimed to develop a suitable statistical model using the associated factors to predict the number of under-five mortality a mother in Nepal encountered throughout her lifetime. The nationally representative Nepal Demographic and Health Survey (NDHS) 2022 data were used to conduct this study. The response variable was the number of under-five mortality a mother has experienced throughout her lifetime. Factors related to different circumstances, including maternal, paternal, socioeconomic, child, environmental, and the utilization of health care services were considered as independent variables. The zero-inflated negative binomial regression (ZINBR) model was fitted as the most plausible model to analyze the number of under-five mortality by its covariates after assessing the different count models including the Poisson regression (PR) model, negative binomial regression (NBR) model, zero-inflated negative binomial regression (ZINBR) model, and hurdle negative binomial regression (HNBR) model. The analysis of the data revealed that 8.6% of the mothers in Nepal have endured at least one under-five mortality. The mean number of deaths before attaining five years was found to be 0.11 (95% CI: 0.10-0.11). The ecological region, smoking habit of mothers, total CEB, source of drinking water, preceding birth interval, and birth order number were obtained as significant covariates of the number of under-five mortality. Mothers from the mountain region (IRR: 1.1756, 95% CI: 1.0095-1.369, p-value = 0.0418) in contrast to hilly region, smoker mothers (IRR: 1.2067, 95% CI: 1.0392-1.4011, p-value = 0.0141) as compared to non-smoker mothers, mothers with total CEB 3 or more (IRR: 1.588, 95% CI: 1.3116-1.9226, p-value = 0.0353) as compared to 2 or less, preceding birth interval less than 24 months (IRR: 1.2061, 95% CI: 1.0788-1.3485, p-value = 0.0018) as compared to 24 months or more, and birth order number 4th or above (IRR: 3.5681, 95% CI: 3.1512-4.0401, p-value < 0.001) as compared to 1st-3rd were associated with the increased incidence of number of under-five mortality. Furthermore, the mothers who drink water from the tube well (IRR: 0.7719, 95% CI: 0.6191-0.9623, p-value = 0.0323) as compared to those who drink water from other sources have a decreased incidence of the number of under-five mortality. The ecological region, mother smokes cigarettes, total CEB, source of drinking water, preceding birth interval, and the birth order number appeared as the important predictors of under-five mortality. The high under-five mortality prevalent in the mountain and terai areas of Nepal presents a significant hurdle in efforts to decrease overall under-five mortality. Thus, improving existing healthcare services, raising the quality of life, and introducing effective health education initiatives are crucial to reducing under-five mortality.

The study focuses on sight distance restrictions on horizontal curves in highway design. Barriers along open highways and walls on tunnels may restrict the available sight distance in the design of horizontal curves. The study aims to finalize a desirable horizontal radius by examining the equilibrium requirement of the centrifugal force vs. horizontal sight distance constraints. The study examines these restrictions for the conventional case when the stopping sight distance lies within the length of the horizontal curve (case 1) and when the stopping sight distance lies outside the length of the horizontal curve (case 2). Also included in this study is a sensitivity analysis when the stopping sight distance is longer than the length of horizontal curve (case 2), based on the ratio between them. For the conventional case (case 1) the minimum horizontal radius that is based on the equilibrium of centrifugal force, is too small for the requirements of stopping sight distance when there are obstructions along the road that intrude on the line of sight. Sensitivity analysis results for case 2 show that the gap between the available horizontal stopping sight distance (for the minimum horizontal radius) vs. the design (demanded) stopping sight distance values and between the minimum horizontal radius vs. the desirable radius can be totally resolved when the ratio between stopping sight distance and the length of horizontal curve equals to 4.5 (for horizontal sightline offset of 4.8 m, i.e. shoulder width of 3.0 m plus half lane width of 1.8 m). Practically, the resulted conditions of the ratio between stopping sight distance and the length of horizontal curve are not feasible in horizontal curve design especially when the horizontal sightline offset equals to 3.0 m. When the limitations of horizontal stopping sight distance are valid and the obstruction height is higher than the maximum, the highway engineer might consider a tradeoff by increasing the horizontal curve radius as well as adjusting the design to case 2.

Most existing freeway crash frequency models analyze overall frequency of crashes. Furthermore, researchers have traditionally used average annual daily traffic (AADT) to represent traffic volume in their models. These two cases are examples of macroscopic crash frequency modeling. Segregating crashes on the basis of type of crash, peak or off-peak traffic conditions, lighting conditions, severity, and pavement condition could provide insight into the specific factors that affect each category. In this study multiple binary categorizations of the crashes were created to identify the factors associated with their frequencies and used geometric characteristics of the freeway and microscopic traffic variables that were based on loop detector data. These categorizations included multiple- and single-vehicle crashes, peak period and off peak period crashes, dry and wet pavement crashes, daytime and dark-hour crashes, and property-damage-only and injury crashes. Models for frequency of each of the two groups of crashes were estimated separately for all five categorizations. To account for correlation between the disturbance terms arising from omitted variables between any two models in a category, seemingly unrelated negative binomial (SUNB) regression was used for simultaneous estimation. SUNB estimation proved to be advantageous for multiple- and single-vehicle crashes and for daytime and dark-hour crashes. Road curvature and presence of on- or off-ramps were found to be the significant factors related to every crash category. Median type and pavement surface type were among other important factors affecting crashes. AADT was significant in most models and the 15-min coefficient of variation of speed was significant for frequency of daytime and peak period crashes. SUNB estimation proved to increase the efficiency of the crash frequency models by accounting for the disturbance correlation, reducing the standard errors, and providing better model fit.

Objectives: The aim of this study was to identify the risk factors associated with number of crime committed by youth (Youth Delinquency) between ages 10-17, using Ordinary Least Square (OLS), Poisson Regression model (PRM), Negative Binomial Regression model (NBRM)&amp; Zero Inflated Negative Binomial (ZINB) with the aim to choose the most appropriate model for the observed count data. Methodology: The data in the study was collected from youth whose mothers enrolled in Philadelphia Collaborative Perinatal Project (CPP). School and delinquency record (between ages 10-17) was obtained by the Centre for studies in Criminology and Criminal Law. Literature search suggest that factors associated with child delinquency can be divided into four main factors as Individual, Family, School and Peer. Therefore we included variables in the analysis accordingly. Result: For OLS scatter plot of residuals versus estimated counts showed definite pattern of heterogeneity (non-constant variance). The likelihood-ratio (LR) test of over dispersion yields the significant p-value, which implied that the outcome variable is overdispersed. The plot of the difference between the actual probabilities and the mean predicted probabilities for each model showed that PRM has poor predictions for low counts (0-2). Conclusion: NBRM and ZINB both performed well, however fit statistics revealed that NBRM has provided more closed predication as compare ZINB.NB modeling techniques provides much more compelling and accurate results instead of basic PRM or those available through simple linear or log-linear modeling techniques

New highway geometric design methods for minimizing vehicular fuel consumption and improving safety

The Poisson regression model is often used as a first model for count data with covariates. Since this model is a GLM with canonical link, regression parameters can be easily fitted using standard software. However the model requires equidispersion, which might not be valid for the data set under consideration. There have been many models proposed in the literature to allow for overdispersion. One such model is the negative binomial regression model. In addition, score tests have been commonly used to detect overdispersion in the data. However these tests do not allow to quantify the effects of overdispersion. In this paper we propose easily interpretable discrepancy measures which allow to quantify the overdispersion effects when comparing a negative binomial regression to Poisson regression. We propose asymptotic $\alpha$-level tests for testing the size of overdispersion effects in terms of the developed discrepancy measures. A graphical display of p-values curves can then be used to allow for an exact quantification of the overdispersion effects. This can lead to a validation of the Poisson regression or a discrimination of the Poisson regression with respect to the negative binomial regression. The proposed asymptotic tests are investigated in small samples using simulation and applied to two examples.

More than 25% of all roadway fatalities in the U.S. are associated with a horizontal curve, and the average crash rate for horizontal curves is about three times that of other types of highway segments. A focus on horizontal curves can prove to be a cost-effective approach to reducing safety issues. Accurate crash prediction models (CPMs) on horizontal curves can help roadway safety practitioners assess and prioritize safety improvements. Although many CPMs have been developed, there are no extant studies that compare different CPMs on a singular, real-world, large-scale, and comprehensive dataset to evaluate their capability for horizontal curve crash prediction. This study critically evaluated commonly used CPMs, including multiple linear regression (MLR), Poisson regression (PR), negative binomial regression (NBR), support vector machine, random forest (RF), and fully connected neural network (FCNN) models, on rural curves extracted from 18,000 centerline miles of Georgia, U.S.’s state-maintained routes and statewide historical crash data set from 2013 to 2021. Results show PR and NBR models outperform MLR by around 6%. Moreover, the FCNN and RF models further improved this performance by around an additional 6% over the PR and NBR models. Overall, machine learning (ML)-based models outperform generalized linear regression models. The results prove ML-based models can be recommended to transportation agencies to forecast horizontal curve crashes more accurately.

The novel coronavirus causing the COVID-19 pandemic and morbidity and mortality around the world has left unprecedented repercussions on all aspects of life. Estimation of the incidence of coronavirus patients will reimburse planning and preparation for new waves. The link between COVID-19 and air pollution data can allow us to estimate COVID cases. In this study, count data models; Poisson regression and negative binomial regressions models and machine learning methods; support vector regression and random forest are used to investigate the link between new COVID-19 cases and the air pollution data in Bangkok, Thailand. The root means square error and means absolute percentage error are used as the criterions to compare the efficiency from each method. The results from the real data showed that the support vector regression outperforms random forest and also Poisson regression and negative binomial regression models.

This paper describes a research project conducted at Madrid Polytechnic University with the objective of refining the negative binomial accident prediction models that had been developed previously for two-lane rural roads in Spain. Cumulative scaled residuals plots were used to identify the value ranges of annual average daily traffic (AADT) where the model over- or underestimated accident frequencies. They were also used to determine whether the calibration sample contained redundant information for some values of the explanatory variables that was detrimental to the model-fitting process. On the basis of the results of these analyses, two approaches were explored to refine the models. First, a random reduction of the sample size was tried to mitigate the effect of redundancies in the statistical information. Second, the sample was stratified, and independent models were fitted for the regions of AADT values where the cumulative residuals plot showed moderate fluctuations. These processes reduced considerably the amount of over- and underestimation of the models, which indicates that in some cases they may be a valid tool to refine accident prediction models and to overcome the lack of flexibility in the functional forms commonly used in multivariate regression modeling.

Sight distance is an important indicator to ensure the safety of drivers, and is also an indispensable evaluation basis in highway safety engineering. In mountainous highways, high slopes and small radius often lead to poor visibility and traffic accidents. Through the combined calculation of horizontal and vertical sections, this paper comprehensively considers the specific sizes of roadside clearance, high slope, as well as the position and height of the driver’s view point and other factors, and it analyzes the limited visibility of the driver in the process of driving right turn. An effective and simplified calculation method based on design data for three dimensional (3D) stopping sight distance (S.S.D.) in high fill sections is proposed. Finally, the S.S.D. inspection of the actual highway, based on design speed and operating speed, is carried out, and the sight distance of the calculated point is judged by comparing the value with the normal value and the calculation result of the horizontal sightline offset. The results show that the method proposed in this paper is consistent with the sight distance results obtained by the horizontal sightline offset method, which indicates the calculation method is accurate and provides a technical reference for S.S.D. evaluation in highway safety engineering.