A comparative study on machine learning based algorithms for prediction of motorcycle crash severity

Motorcycle crash severity is under-researched in Ghana. Thus, the probable risk factors and association between these factors and motorcycle crash severity outcomes is not known. Traditional statistical models have intrinsic assumptions and pre-defined correlations that, if flouted, can generate inaccurate results. In this study, machine learning based algorithms were employed to predict and classify motorcycle crash severity. Machine learning based techniques are non-parametric models without the presumption of relationships between endogenous and exogenous variables. The main aim of this research is to evaluate and compare different approaches to modeling motorcycle crash severity as well as investigating the effect of risk factors on the injury outcomes of motorcycle crashes. Motorcycle crash dataset between 2011 and 2015 was extracted from the National Road Traffic Crash Database at the Building and Road Research Institute (BRRI) in Ghana. The dataset was classified into four injury severity categories: fatal, hospitalized, injured, and damage-only. Three machine learning based models were developed: J48 Decision Tree Classifier, Random Forest (RF) and Instance-Based learning with parameter k (IBk) were employed to model the severity of injury in a motorcycle crash. These machine learning algorithms were validated using 10-fold cross-validation technique. The three machine learning based algorithms were compared with one another and the statistical model: multinomial logit model (MNLM). Also, the relative importance analysis of the attribute was conducted to determine the impact of these attributes on injury severity outcomes. The results of the study reveal that the predictions of machine learning algorithms are superior to the MNLM in accuracy and effectiveness, and the RF-based algorithms show the overall best agreement with the experimental data out of the three machine learning algorithms, for its global optimization and extrapolation ability. Location type, time of the crash, settlement type, collision partner, collision type, road separation, road surface type, the day of the week, and road shoulder condition were found as the critical determinants of motorcycle crash injury severity.

[1]  Dominique Lord,et al.  The statistical analysis of highway crash-injury severities: a review and assessment of methodological alternatives. , 2011, Accident; analysis and prevention.

[2]  F. Afukaar,et al.  Prevalence of Helmet Use Among Motorcycle Users in Tamale Metropolis, Ghana: An Observational Study , 2010, Traffic injury prevention.

[3]  D. Kibler,et al.  Instance-based learning algorithms , 2004, Machine Learning.

[4]  Deo Chimba,et al.  The Prediction of Highway Traffic Accident Injury Severity with Neuromorphic Techniques , 2009 .

[5]  Ryan Doczy,et al.  Machine Learning Methods to Analyze Injury Severity of Drivers from Different Age and Gender Groups , 2018, Transportation Research Record: Journal of the Transportation Research Board.

[6]  George Yannis,et al.  Overview of critical risk factors in Power-Two-Wheeler safety. , 2012, Accident; analysis and prevention.

[7]  Konstantina Gkritza,et al.  A mixed logit analysis of two-vehicle crash severities involving a motorcycle. , 2013, Accident; analysis and prevention.

[8]  Alexander Maistros,et al.  A comparison of contributing factors between alcohol related single vehicle motorcycle and car crashes. , 2014, Journal of safety research.

[9]  Seyed Mohammad Hossein Hasheminejad,et al.  Traffic accident severity prediction using a novel multi-objective genetic algorithm , 2017 .

[10]  Mohamed Abdel-Aty,et al.  Examining traffic crash injury severity at unsignalized intersections. , 2010, Journal of safety research.

[11]  Tai-Jin Song,et al.  Injury severity in delivery-motorcycle to vehicle crashes in the Seoul metropolitan area. , 2014, Accident; analysis and prevention.

[12]  Characteristics of Pedestrian Accidents on Trunk Roads in Ghana , 2013 .

[13]  Chao Wang,et al.  Road Traffic Congestion and Crash Severity: Econometric Analysis Using Ordered Response Models , 2010 .

[14]  Anael Sam,et al.  Diabetes Forecasting Using Supervised Learning Techniques , 2014 .

[15]  Hoong Chor Chin,et al.  An analysis of motorcycle injury and vehicle damage severity using ordered probit models. , 2002, Journal of safety research.

[16]  Fang Liu,et al.  An Improved Fuzzy Neural Network for Traffic Speed Prediction Considering Periodic Characteristic , 2017, IEEE Transactions on Intelligent Transportation Systems.

[17]  Matthieu de Lapparent,et al.  Empirical Bayesian analysis of accident severity for motorcyclists in large French urban areas. , 2006 .

[18]  Francis K Afukaar,et al.  Speed control in developing countries: issues, challenges and opportunities in reducing road traffic injuries , 2003, Injury control and safety promotion.

[19]  Yanyong Guo,et al.  Exploring unobserved heterogeneity in bicyclists' red-light running behaviors at different crossing facilities. , 2018, Accident; analysis and prevention.

[20]  Sachin Kumar,et al.  Severity analysis of powered two wheeler traffic accidents in Uttarakhand, India , 2017 .

[21]  Yoonjin Yoon,et al.  Evaluation of motorcycle safety strategies using the severity of injuries. , 2013, Accident; analysis and prevention.

[22]  J. Ross Quinlan,et al.  C4.5: Programs for Machine Learning , 1992 .

[23]  Amirfarrokh Iranitalab,et al.  Comparison of four statistical and machine learning methods for crash severity prediction. , 2017, Accident; analysis and prevention.

[24]  Zhibin Li,et al.  Identifying the factors affecting bike-sharing usage and degree of satisfaction in Ningbo, China , 2017, PloS one.

[25]  Chih-Wei Pai,et al.  Motorcyclists violating hook-turn area at intersections in Taiwan: an observational study. , 2013, Accident; analysis and prevention.

[26]  Yanyong Guo,et al.  Exploring Evasive Action–Based Indicators for PTW Conflicts in Shared Traffic Facility Environments , 2018, Journal of Transportation Engineering, Part A: Systems.

[27]  Grazia La Cava,et al.  A logistic model for Powered Two-Wheelers crash in Italy , 2012 .

[28]  Li-Yen Chang,et al.  Data mining of tree-based models to analyze freeway accident frequency. , 2005, Journal of safety research.

[29]  Hoong Chor Chin,et al.  Modeling fault among motorcyclists involved in crashes. , 2009, Accident; analysis and prevention.

[30]  J. Guzmán.  Regression Models for Categorical Dependent Variables Using Stata , 2013 .

[31]  Carol A C Flannagan,et al.  Identification and validation of a logistic regression model for predicting serious injuries associated with motor vehicle crashes. , 2011, Accident; analysis and prevention.

[32]  Li-Yen Chang,et al.  Analysis of driver injury severity in truck-involved accidents using a non-parametric classification tree model , 2013 .

[33]  P. Kola Sujatha,et al.  Performance Evaluation Of Classifiers For Analysis Of Road Accidents , 2017, 2017 Ninth International Conference on Advanced Computing (ICoAC).

[34]  Chandra R. Bhat,et al.  Unobserved heterogeneity and the statistical analysis of highway accident data , 2016 .

[35]  Chih-Wei Pai,et al.  A mixed logit analysis of motorists' right-of-way violation in motorcycle accidents at priority T-junctions. , 2009, Accident; analysis and prevention.

[36]  Fan Ye,et al.  Comparing Three Commonly Used Crash Severity Models on Sample Size Requirements : Multinomial Logit , Ordered Probit and Mixed Logit Models , 2013 .

[37]  Yinhai Wang,et al.  Short-Term Speed Prediction Using Remote Microwave Sensor Data: Machine Learning versus Statistical Model , 2016 .

[38]  Hoong Chor Chin,et al.  Severity of driver injury and vehicle damage in traffic crashes at intersections: a Bayesian hierarchical analysis. , 2008, Accident; analysis and prevention.

[39]  George Yannis,et al.  A review of powered-two-wheeler behaviour and safety , 2015, International journal of injury control and safety promotion.

[40]  Richard Amoh-Gyimah,et al.  The effect of natural and built environmental characteristics on pedestrian-vehicle crash severity in Ghana , 2017, International journal of injury control and safety promotion.

[41]  Millicent Awialie Akaateba,et al.  Correlates and Barriers Associated with Motorcycle Helmet Use in Wa, Ghana , 2015, Traffic injury prevention.

[42]  Melissa A Clark,et al.  Correlates of motorcycle helmet use among recent graduates of a motorcycle training course. , 2010, Accident; analysis and prevention.

[43]  Leo Breiman,et al.  Random Forests , 2001, Machine Learning.

[44]  Matthieu de Lapparent,et al.  Empirical Bayesian analysis of accident severity for motorcyclists in large French urban areas. , 2006, Accident; analysis and prevention.

[45]  Konstantina Gkritza,et al.  Differences in Motorcycle Conspicuity-related Factors and Motorcycle Crash Severities in Daylight and Dark Conditions , 2011 .

[46]  Wafaa Saleh,et al.  An analysis of motorcyclist injury severity under various traffic control measures at three-legged junctions in the UK , 2007 .

[47]  Leili Abedi,et al.  Epidemiological pattern of motorcycle injuries with focus on riding purpose: Experience from a middle-income country , 2015 .

[48]  Shan Suthaharan,et al.  Machine Learning Models and Algorithms for Big Data Classification , 2016 .

[49]  Mohammad Mehdi Besharati,et al.  A data mining approach to investigate the factors influencing the crash severity of motorcycle pillion passengers. , 2014, Journal of safety research.

[50]  Jinjun Tang,et al.  Crash injury severity analysis using a two-layer Stacking framework. , 2019, Accident; analysis and prevention.

[51]  Shakil Mohammad Rifaat,et al.  Severity of motorcycle crashes in Calgary. , 2012, Accident; analysis and prevention.

[52]  P. Pooja,et al.  Identification of Cardiac Arrhythmias using ECG , 2012 .

[53]  Li-Yen Chang,et al.  Analysis of traffic injury severity: an application of non-parametric classification tree techniques. , 2006, Accident; analysis and prevention.

[54]  Y. Zou,et al.  A semi-nonparametric Poisson regression model for analyzing motor vehicle crash data , 2018, PloS one.

[55]  Nataliya V Malyshkina,et al.  Markov switching multinomial logit model: An application to accident-injury severities. , 2008, Accident; analysis and prevention.

[56]  Fred Mannering,et al.  Probabilistic models of motorcyclists' injury severities in single- and multi-vehicle crashes. , 2007, Accident; analysis and prevention.

[57]  Ali Tavakoli Kashani,et al.  Identifying the Most Important Factors in the At-Fault Probability of Motorcyclists by Data Mining, Based on Classification Tree Models , 2017, International Journal of Civil Engineering.

[58]  Sunil Patil,et al.  Analysis of Motorcycle Crashes in Texas with Multinomial Logit Model , 2011 .

[59]  Shaibu Bawa,et al.  Prevalence rate of helmet use among motorcycle riders in Kumasi, Ghana , 2018, Traffic injury prevention.

[60]  Ross A Blackman,et al.  Comparison of moped, scooter and motorcycle crash risk and crash severity. , 2013, Accident; analysis and prevention.

[61]  Chih Wei Pai,et al.  Motorcyclist injury severity in angle crashes at T-junctions: Identifying significant factors and analysing what made motorists fail to yield to motorcycles , 2009 .

[62]  James Damsere-Derry,et al.  Road accident fatality risks for “vulnerable” versus “protected” road users in northern Ghana , 2017, Traffic injury prevention.

[63]  Juan de Oña,et al.  Injury severity models for motor vehicle accidents: a review , 2013 .

[64]  Yajie Zou,et al.  Empirical Bayes estimates of finite mixture of negative binomial regression models and its application to highway safety , 2018 .

[65]  J. Awoonor-Williams,et al.  Economic burden of motorcycle accidents in Northern Ghana. , 2011, Ghana medical journal.

[66]  Yanyong Guo,et al.  Evaluation of Factors Affecting E-Bike Involved Crash and E-Bike License Plate Use in China Using a Bivariate Probit Model , 2017 .

[67]  Tom Fawcett,et al.  An introduction to ROC analysis , 2006, Pattern Recognit. Lett..

[68]  Millicent Awialie Akaateba,et al.  A cross-sectional observational study of helmet use among motorcyclists in Wa, Ghana. , 2014, Accident; analysis and prevention.

[69]  R. Dinye.  The significance and issues of motorcycle transport in the Urban areas in northern Ghana , 2013 .

[70]  Richard Amoh-Gyimah,et al.  The effect of road and environmental characteristics on pedestrian hit-and-run accidents in Ghana. , 2013, Accident; analysis and prevention.

[71]  F P Rivara,et al.  Epidemiology of transport-related injuries in Ghana. , 1999, Accident; analysis and prevention.

[72]  Chandra R. Bhat,et al.  Analytic methods in accident research: Methodological frontier and future directions , 2014 .

[73]  Williams Ackaah,et al.  Analysis of fatal road traffic crashes in Ghana , 2011, International journal of injury control and safety promotion.

[74]  Xiaoyu Zhu,et al.  A comprehensive analysis of factors influencing the injury severity of large-truck crashes. , 2011, Accident; analysis and prevention.

[75]  Francis K Afukaar,et al.  Pattern of road traffic injuries in Ghana: Implications for control , 2003, Injury control and safety promotion.

[76]  Tarek Sayed,et al.  Evaluating the safety impacts of powered two wheelers on a shared roadway in China using automated video analysis , 2018 .

[77]  C. Dolea,et al.  World Health Organization , 1949, International Organization.

[78]  Mohamed Abdel-Aty,et al.  A classification tree based modeling approach for segment related crashes on multilane highways. , 2010, Journal of safety research.

[79]  Juan de Oña,et al.  Analysis of traffic accident severity using Decision Rules via Decision Trees , 2013, Expert Syst. Appl..

[80]  Antonio D’Ambrosio,et al.  Analysis of powered two-wheeler crashes in Italy by classification trees and rules discovery. , 2012, Accident; analysis and prevention.

[81]  E. Lagarde.  Road Traffic Injury Is an Escalating Burden in Africa and Deserves Proportionate Research Efforts , 2007, PLoS medicine.