Data mining of tree-based models to analyze freeway accident frequency

INTRODUCTION: Statistical models, such as Poisson or negative binomial regression models, have been employed to analyze vehicle accident frequency for many years. However, these models have their own model assumptions and pre-defined underlying relationship between dependent and independent variables. If these assumptions are violated, the model could lead to erroneous estimation of accident likelihood. Classification and Regression Tree (CART), one of the most widely applied data mining techniques, has been commonly employed in business administration, industry, and engineering. CART does not require any pre-defined underlying relationship between target (dependent) variable and predictors (independent variables) and has been shown to be a powerful tool, particularly for dealing with prediction and classification problems. METHOD: This study collected the 2001-2002 accident data of National Freeway 1 in Taiwan. A CART model and a negative binomial regression model were developed to establish the empirical relationship between traffic accidents and highway geometric variables, traffic characteristics, and environmental factors. RESULTS: The CART findings indicated that the average daily traffic volume and precipitation variables were the key determinants for freeway accident frequencies. By comparing the prediction performance between the CART and the negative binomial regression models, this study demonstrates that CART is a good alternative method for analyzing freeway accident frequencies. IMPACT ON INDUSTRY: By comparing the prediction performance between the CART and the negative binomial regression models, this study demonstrates that CART is a good alternative method for analyzing freeway accident frequencies.     

Random parameter models for accident prediction on two-lane undivided highways in India

Introduction

Generalized linear modeling (GLM), with the assumption of Poisson or negative binomial error structure, has been widely employed in road accident modeling. A number of explanatory variables related to traffic, road geometry, and environment that contribute to accident occurrence have been identified and accident prediction models have been proposed. The accident prediction models reported in literature largely employ the fixed parameter modeling approach, where the magnitude of influence of an explanatory variable is considered to be fixed for any observation in the population. Similar models have been proposed for Indian highways too, which include additional variables representing traffic composition. The mixed traffic on Indian highways comes with a lot of variability within, ranging from difference in vehicle types to variability in driver behavior. This could result in variability in the effect of explanatory variables on accidents across locations. Random parameter models, which can capture some of such variability, are expected to be more appropriate for the Indian situation.

Method

The present study is an attempt to employ random parameter modeling for accident prediction on two-lane undivided rural highways in India. Three years of accident history, from nearly 200 km of highway segments, is used to calibrate and validate the models.

Results

The results of the analysis suggest that the model coefficients for traffic volume, proportion of cars, motorized two-wheelers and trucks in traffic, and driveway density and horizontal and vertical curvatures are randomly distributed across locations.

Conclusions

The paper is concluded with a discussion on modeling results and the limitations of the present study.     

Application of generalized link functions in developing accident prediction models

The Bayesian Poisson–Gamma hierarchy, leading to the negative binomial distribution, has been the standard practice in developing accident prediction models. To linearize the relationship connecting the mean of the negative binomial distribution to relevant covariates, a canonical log link has traditionally been used. Typically, little information is available regarding the choice of a particular link. To avoid link misspecification, it is proposed to nest the canonical log link model within a generalized link family and subsequently use the full Bayes method for parameter estimation, performance evaluation and inference. The proposed approach was applied to a sample of accident and traffic volume data corresponding to 99 intersections in the city of Edmonton, Alberta. The results showed that both the generalized link model and the traditional canonical link model provided adequate fit to the data. However, the Bayes factor provided a clear statistical support for the use of the generalized link approach. A procedure for link validation is also described. It allows the users (e.g., road authorities) to consider the changes in predicted accidents that will result if a generalized link is used instead of a canonical link. If a certain maximal change is tolerated, the canonical link can be used to analyze the data; otherwise the generalized link is worth the extra efforts and should be adopted. When compared with the traditional approach, the generalized link model was found to predict a lower number of accidents whenever there is a heavy traffic at the major approach, especially if combined with light flow on the minor approach. The paper concludes by identifying out areas for further research.     

Time series analysis of coal mine accident experience

This study investigates several forecasting techniques that can be useful to mine safety managers for studying mine accident rate behavior. Three time series models were studied for extrapolation of accident rates. These models are applied to historical accident incidence data from a coal mine. Further, a method is presented for evaluating the three models for the selection of an appropriate model. For this particular mine application, it is concluded that the more complex Box-Jenkins ARMA model as well as first order autoregressive model do not give better results than the simple exponential smoothing model. However, when the random variations or autocorrelations in the accident experience rates between periods are different, the models may predict differently. As such, specific models must be developed for each mine on the basis of statistical analysis of the mine accident experience data over time. Moreover, the importance of incorporating human judgement to interpret the results of statistical forecasting cannot be overemphasized. Integration of policy or operating changes, which may impact mine safety performance, with statistical forecasting techniques is essential to arrive at a realistic prediction of future performance.     

航空装备事故的灰色时序组合预测模型

为提高航空装备事故预测水平,提出一种基于灰色和时间序列分析模型的航空装备事故组合预测模型。先构建灰色模型,提取历史数据中承载的趋势信息。然后进行模型选择、阶数识别和参数估计,建立灰色残差的时间序列分析模型,用以刻画历史数据中的随机波动特征。最后,将2个模型的预测值相加,得到所求的组合预测结果。实例中,以美国空军1996—1999年的A级飞行事故10万时率数据为基础,建立灰色时序组合模型,模型中短期预测精度优于单一灰色模型,平均相对误差控制在5%以内,预测结果能够反映航空装备安全的实际状况。     

Road casualties and enforcement: distributional assumptions of serially correlated count data

OBJECTIVE: Road safety data are often in the form of counts and usually temporally correlated. The objective of this research is to investigate the distributional assumptions of road safety data in the presence of temporal correlation. METHODS: Using the generalized linear model framework, four distributional assumptions are considered: normal, Poisson, quasi-Poisson and negative binomial, and appropriate models are estimated. Monthly casualty and police enforcement data from Greece for a period of six years (January 1998-December 2003) have been used. The developed models include sinusoidal latent terms to capture the temporal serial correlation of observations. Several statistical goodness-of-fit diagnostic tests have been performed for the results of the estimated models, and the predictive capabilities of the models are investigated. RESULTS: The residuals of the quasi-Poisson and negative binomial models do not show any serial correlation. The signs of the estimated coefficients for all models are consistent and intuitive. In particular, a negative coefficient value for the number of breath alcohol controls indicates that the number of persons killed and seriously injured decreases as the intensity of breath alcohol controls increases. The Poisson model fails to capture the overdispersion in the data, thus underestimating the standard errors of the estimated coefficients. CONCLUSION: The results suggest that the quasi-Poisson and negative binomial outperform the normal and Poisson models in this application. The findings of this research demonstrate a clear link between the intensification of police enforcement and the reduction of traffic accident casualties. In particular, an increase in the number of breath alcohol controls in Greece after 1998 contributed to a reduction in the number of persons killed and seriously injured from traffic accidents.     

基于车速的交通事故贝叶斯预测

为了降低交通事故的发生率 ,提高道路交通安全水平 ,提出了基于车速的贝叶斯预测方法来检测和预测交通事故。首先对车速与交通事故之间的关系进行分析。在分析的基础上 ,以车速为衡量对象 ,提出贝叶斯预测方法。通过使用车速观测数据 ,应用 χ2 检验 ,确定是否为异常数据 ;并通过最小风险的贝叶斯预测 ,确定该异常是否会导致交通事故。最后 ,绘出利用该贝叶斯预测方法进行交通事故预测的流程图     

关于铁路行车事故预测的应用研究

本文论述了事故预测对铁路行车安全的重要性,并应用灰色预测模型和概率回归估计法模型研制了预测软件,进行铁路行车事故的宏观及微观预测,为采取有针对性的防范措施提供一个可靠信息。     

道路交通事故发展趋势分析与预测

根据福建省2000 -2010年交通事故相关指标,采用统计图表分析法进行交通事故发展趋势分析与安全水平比较研究,结果表明交通事故各项绝对指标总体呈下降趋势,但从万车死亡率、受伤人数与死亡人数比及交通事故死亡人数占各类事故死亡人数比重等相对指标看,交通安全总体水平偏低,交通事故后果比较严重.对交通事故死亡人数与GDP、机动车保有量、公路通车里程、人口数四项影响因素进行了多元线性回归分析,分析得出四个影响因素总体对交通事故死亡人数的线性影响是显著的,采取向后筛选策略线性回归分析得出,死亡人数与GDP的线性关系是显著的,根据回归结果建立了交通事故的预测模型.     

应用改进的数量化理论预测道路断面事故率

为准确预测道路断面的安全性,建立道路交通事故断面事故率模型。首先选定事故率作为微观预测目标,选取驾驶员的驾龄、车道数、平曲线半径、纵坡度、路面情况、路口路段类型、道路宽度7项因素作为主要影响因素,并且将各影响因素分为若干类目。在数量化理论的基础之上建立改进的数量化理论的道路断面事故率预测模型,最后以某国道222.888～377.387 km段作为算例进行计算,并选取事故多发段333.5～334.0 km处对模型进行具体应用。研究结果表明,对事故影响最大的是该路段中三枝交叉口,其次是3 a(含3 a)以下驾龄及四枝交叉口。     

我国双车道公路事故预测模型研究中数据采集

事故预测模型的研究可为公路设计人员提供有效的安全设计工具。该研究需要大量基础数据的支撑,数据采集量的大小、数据质量和数据分析水平直接决定研究成果的优劣。笔者在双车道公路事故预测模型研究过程中,对我国双车道公路安全影响要素进行分析,确定基于现有条件下比较理想的数据采集对象,总结了数据采集的方法及其不足之处。具体包括数据采集的范围,样本量的需求,事故数据、交通组成数据、路侧、接入口和路线数据采集的方法、数据处理方式等。     

事故预测数学模型的研究与实践

依据收集统计的、真实可靠的事故数据,运用数学方法及建立的多种事故预测模型,在工作中理论与实际相结合,提出并建立了事故预测数学模型,给出具体方法与步骤,以及应用中应遵循的原则和规律,从而做到对将来发生的事故未卜先知,用以指导安全生产活动,预防事故发生。同时,根据我国近18年来各类事故死亡人员的数据建立事故预测数学模型,并运用确立的最佳事故预测数学模型而预测出的事故数据,探讨我国伤亡事故发生的趋势。     

灰理论在企业职工工伤事故频率预测中的应用

简述了灰理论中GM(1,1)动态预测模型的特点及其建模过程,并将其应用于企业职工工伤事故频率预测中。经检验,模型的精度良好,预测结果可靠,有利于企业发现和掌握安全生产规律,以及对生产的未来状态作出科学的定量预测。通过实例验证了该模型的有效性。     

安全投资与安全效益的预测及关联分析

运用灰色预测模型,建立了安全投资与事故损失预测模型.运用灰色关联理论,建立了安全投资与安全效益灰色关联分析模型,并利用这2个模型对某石化公司的安全分项投资及事故直接经济损失情况进行了案例分析.初步确定该企业安全投资中各分项投资与事故损失的预测值.确定了该企业安全投资中各分项投资与效益的相关程度,找出了安全投资结构中与效益关联度最大的因素,并优先将其作为安全投资方向.     

水上交通事故分析研究进展

为更好地开展水上交通事故分析研究,提高我国水上交通安全水平,从船舶风险评估与事故预测、事故分析以及事故及通航安全数据的组织与数据库建立3个方面对国内外的相关研究进行论述和分析。提出以建立水上交通事故时空数据平台为基础,结合数据挖掘和安全工程的理论方法研究事故发生机理,评估事故风险,并将研究事故模拟再现技术作为事故分析的重要技术手段。     

Systemic accident risk assessment in air traffic by Monte Carlo simulation

A systemic accident model considers accidents as emergent phenomena from variability and interactions in a complex system. Air traffic risk assessments have predominantly been done by sequential and epidemiological accident models. In this paper we demonstrate that Monte Carlo simulation of safety relevant air traffic scenarios is a viable approach for systemic accident assessment. The Monte Carlo simulations are based on dynamic multi-agent models, which represent the distributed and dynamic interactions of various human operators and technical systems in a safety relevant scenario. The approach is illustrated for a particular runway incursion scenario, which addresses an aircraft taxiing towards the crossing of an active runway while its crew has inappropriate situation awareness. An assessment of the risk of a collision between the aircraft taxiing with an aircraft taking-off is presented, which is based on dedicated Monte Carlo simulations in combination with a validation approach of the simulation results. The assessment particularly focuses on the effectiveness of a runway incursion alert system that warns an air traffic controller, in reducing the safety risk for good and reduced visibility conditions.     

Investigating the different characteristics of weekday and weekend crashes

IntroductionThis study provides a systematic approach to investigate the different characteristics of weekday and weekend crashes.MethodWeekend crashes were defined as crashes occurring between Friday 9 p.m. and Sunday 9 p.m., while the other crashes were labeled as weekday crashes. In order to reveal the various features for weekday and weekend crashes, multi-level traffic safety analyses have been conducted. For the aggregate analysis, crash frequency models have been developed through Bayesian inference technique; correlation effects of weekday and weekend crash frequencies have been accounted. A multivariate Poisson model and correlated random effects Poisson model were estimated; model goodness-of-fits have been compared through DIC values. In addition to the safety performance functions, a disaggregate crash time propensity model was calibrated with Bayesian logistic regression model. Moreover, in order to account for the cross-section unobserved heterogeneity, random effects Bayesian logistic regression model was employed.ResultsIt was concluded that weekday crashes are more probable to happen during congested sections, while the weekend crashes mostly occur under free flow conditions. Finally, for the purpose of confirming the aforementioned conclusions, real-time crash prediction models have been developed. Random effects Bayesian logistic regression models incorporating the microscopic traffic data were developed. Results of the real-time crash prediction models are consistent with the crash time propensity analysis. Furthermore, results from these models would shed some lights on future geometric improvements and traffic management strategies to improve traffic safety.Impact on IndustryUtilizing safety performance to identify potential geometric improvements to reduce crash occurrence and monitoring real-time crash risks to pro-actively improve traffic safety.     

基于Hankel-DMD的城市交通事故风险时空预测*

为解决城市交通事故风险时空分布预测任务中时空关联性捕捉困难的问题,提出基于动态模态分解（DMD）的城市交通事故分析时空预测模型,模型利用总最小二乘法去除交通事故数据中的噪声,应用结合Hankel矩阵的动态模态分解模型（Hankel-DMD）捕捉交通事故风险的时空关联性,对交通事故风险的时空分布进行预测。研究结果表明:DMD框架能够为高维预测任务提供低秩解决方案,从高维数据中捕捉时空关联性;Hankel-DMD模型在预测评价指标平均绝对误差和均方根误差方面的表现明显优于统计学及机器学习等方法;Hankel-DMD模型产生的动态模态和特征值,对事故风险系统的时空动态特征具有一定的可解释性,同时验证Hankel-DMD模型的适用性。     

道路交通安全研究方法

介绍了交通安全的研究方法和预测模型 ,并对统计分析、模糊数学分析及灰色理论方法进行了比较。在此基础上 ,介绍了适应我国目前交通安全研究现状的交通事故灰色预测研究方法 ,对交通事故发展趋势进行了预测。论文给出了相应的灰色预测模型及预测方法 ,预测结果与实际情况较吻合     

Safety performance functions with measurement errors in traffic volume

One of the most common and important predictors in safety performance functions (SPFs) is traffic volume which is known to be measured with uncertainty. Such measurement errors (ME) can attenuate the respective predictors’ effect and also increase dispersion. This paper proposes an approach which involves the use of a ME model based on traffic flow time series data. The model is used in conjunction with the negative binomial SPF to circumvent the bias in predicting the aggregate number of accidents during the time period under study. The proposed approach (denoted by MENB), was compared with the traditional negative binomial (NB) technique by way of Monte Carlo simulation. Furthermore, both approaches were applied to two datasets corresponding to 131 and 130 road segments in British Columbia. The full Bayes method was utilized for parameter estimation, performance evaluation and inference through the use of Markov Chain Monte Carlo (MCMC) techniques. The simulation results showed that MENB has outperformed NB when large measurement errors are present. The goodness-of-fit statistics showed that MENB has provided a slightly better fit to the data. However, in the presence of measurement errors, the NB has underestimated the predicted number of accidents for heavy traffic on long road segments and vice versa. The use of MENB is justified when the variance in volume between years is large otherwise both approaches yield comparable results.