Freeway Incident Frequency Analysis Based on CART Method
Abstract
Classification and Regression Tree (CART), one of the most widely applied data mining techniques, is based on the classification and regression model produced by binary tree structure. Based on CART method, this paper establishes the relationship between freeway incident frequency and roadway characteristics, traffic variables and environmental factors. The results of CART method indicate that the impact of influencing factors (weather, weekday/weekend, traffic flow and roadway characteristics) of incident frequency is not consistent for different incident types during different time periods. By comparing with Negative Binomial Regression model, CART method is demonstrated to be a good alternative method for analyzing incident frequency. Then the discussion about the relationship between incident frequency and influencing factors is provided, and the future research orientation is pointed out.
References
Lindley JA. Urban freeway congestion: quantification of the problem and effectiveness of potential solutions. ITE J. 1987 Jan;57:27-32.
Washington SP, Karlaftis MG, Mannering FL. Statistical and econometric methods for transportation data Analysis, New York: Chapman and Hall/CRC; 2003.
Breiman L, Friedman JH, Olshen RA, Stone CJ. Classification and regression trees. London: Chapman & Hall/CRC; 1998.
Chang LY, Chen WC. Data mining of tree-based models to analyze freeway accident frequency. J Safety Res. 2005;36(4):365-375.
Yang P, Wu B. Traffic management and control. Beijing: Renmin Traffic Press; 2004.
Lord D, Mannering FL. The statistical analysis of crash-frequency data: a review and assessment of methodological alternatives. Transp Res Part A Policy Pract. 2010 Jun;44(5):291–305.
Poch M, Mannering FL. Negative binomial analysis of intersection-accident frequencies. J Transp Eng. 1996 Mar;122(2):105–113.
Ivan JN, Wang C, Bernardo NR. Explaining two-lane highway crash rates using land use and hourly exposure. Accid Anal Prev. 2000 Nov;32(6):787–795.
Carson J, Mannering FL. The effect of ice warning signs on accident frequencies and severities. Accid Anal Prev. 2001 Jan;33(1):99-109.
Shankar VN. Limited dependent variable and structural equations models: empirical applications to traffic operations and safety. Dissertation: University of Washington; 1997.
Karlaftis MG, Tarko AP. Heterogeneity considerations in accident modeling. Accid Anal Prev. 1998 Jul;30(4):425-433.
Kuhnert PM, Do K, McClure R. Combining non-parametric models with logistic regression: an application to motor vehicle injury data. Comput Stat Data Anal. 2000 Sept;34(3):371-386.
Karlaftis MG., Golias I. Effects of road geometry and traffic volumes on rural roadway accident rates. Accid Anal Prev, 2002 May;34(3):357-365.
Chang LY, Wang HW. Analysis of traffic injury severity: an application of non-parametric classification tree techniques. Accid Anal Prev, 2006 Sept;38(5):1019-1027.
Kashani AT, Mohaymany AS. Analysis of the traffic injury severity on two-lane, two-way rural roads based on classification tree models. Safety Sci. 2011 Dec;49(10):1314-1320.
Pakgohar A, Tabrizi RS, Khalili M, Esmaeili A. The role of human factors in incidence and severity of road crashes based on CART and LR regression: a data mining approach. Procedia Comput Sci. 2011;3:764-769.
Yap BW, Norashikin N, Wong, SV, Mohamad AL. Decision tree model for count data. Proceedings of the World Congress on Engineering 2012. Vol I; July 4-6, 2012, London, U.K.
Anastasopoulos PCh, Mannering FL. A note on modeling vehicle-accident frequencies with random parameter count models. Accid Anal Prev. 2009 Jan;41(1):153–159.
El-Basyouny K, Sayed T. Accident prediction models with random corridor parameters. Accid Anal Prev. 2009 Sept;41(5):1118–1123.
Wang X, Abdel-Aty M. Temporal and spatial analyses of rear-end crashes at signalized intersections. Accid Anal Prev. 2006 Nov;38(6):1137–1150.
Xu X, Kwigizile V, Teng H. Identifying access management factors associated with safety of urban arterials mid-blocks: a panel data simultaneous equation models approach. Traffic Inj Prev. 2013;14(7):734-742.
Authors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).
