Pemodelan Data Kecelakaan Lalu Lintas Menggunakan Metode Regresi Poisson dan Regresi Binomial Negatif

Mochamad Syaeful Fuad, Abdul Kudus, Siti Sunendiari

Abstract


Abstract. Road accidents in Bandung City seems to be on ascendency and the root causes have been attributed to issues such as human errors and superstitions. Since the occurrences of these accidents are discrete, they are often modelled using regression models. It is therefore the purpose of this study to determine an appropriate regression model that adequately fits road accidents in Bandung City. Several models were compared to fit count data that encounter the field of transportation. These models include Poisson and Negative Binomial (NB) models. In order to compare the performance of these models, the various model selection methods such as Deviance goodness of fit test, Akaike’s Information Criterion (AIC) were employed. Based on road accidents  data in Bandung City from January 2014 until December 2016, the values of Deviance goodness of fit test and AIC for the Poisson model was the smallest as compared to that of the NB models, it appeared that, the Poisson model performed best than the Negative Binomial (NB) models. Base on the model selected (Poisson model), the predictors that contributed significantly and also had a high effect on the expected or mean number of persons killed in road accidents within a particular period were Collision type, Loss of control as Driver errors,  and Type of vehicle.

Keywords : AIC, Goodness Of Fit, Poisson Model, Negative Binomial Model.

 

Abstrak. Kecelakaan  lalu lintas di kota Bandung nampaknya berpengaruh dan akar penyebabnya disebabkan oleh isu-isu seperti kesalahan manusia dan takhayul. Karena kejadian kecelakaan ini bersifat diskrit, banyak yang memodelkan dengan menggunakan model regresi. Oleh karena itu tujuan penelitian ini untuk mengetahui model regresi yang tepat dan cocok untuk data kecelakaan lalu lintas di Kota Bandung. Model yang akan dibandingkan yaitu Model Poisson dan Binomial Negatif. Untuk membandingkan kinerja model ini, maka digunakanlah metode pemilihan model goodness of fit dan Akaike Information Criteria (AIC). Berdasarkan data kecelakaan lalu lintas Kota Bandung dari Januari 2014 sampai Desember 2016, nilai AIC untuk model Poisson lebih kecil dibandingkan Binomial Negatif, maka Model Poisson dinilai lebih baik dari pada Model Binomial Negatif. Berdasarkan model yang terpilih (Model Poisson), prediktor yang memberikan kontribusi signifikan dan juga memiliki efek yang tinggi terhadap jumlah orang yang meninggal dalam kecelakaan lalu lintas dalam periode tertentu adalah jenis kecelakaan, faktor pengemudi dan jenis kendaraan yang terlibat.

Kata kunci :  AIC, Goodness Of Fit, Model Poisson, Model Binomial Negatif.


Keywords


AIC, Goodness Of Fit, Model Poisson, Model Binomial Negatif.

Full Text:

PDF

References


Abdel-Aty, M.A., Radwan, A.E. 2000, Modelling traffic accidents occurrence and involvement. Accident Analysis and Prevention, 32, 633-642. (http://dx.doi.org/10.1016-4575(99)00094-9).

Amanda, Gita. (2014). Indonesia Urutan Pertama Peningkatan Kecelakaan Lalu Lintas. Republika Online. (http://www.republika.co.id/berita/nasional/umum/4/11/06/nem9nc-indonesia-urutan-pertama-peningkatan-kecelakaan-lalu-lintas. Diakses pada 10 September 2017)

Fletcher J.P., Baguley C.J., Sexton B., and Done S. 2006. Road accident modeling for highway development countries. Main Report Trials in India and Tanzania. Report No. PPRO95.

Hilbe M, Joseph. (2011). Negative Binomial Regression. Second Edition. 188- 191

Joshua, S.C., Garber, N.J., 1990. Estimating truck accident rate and involvements using linear and Poisson regression models. Transportation Planning and Technology, 15(1), 41-58.

Jovanis, P.P., and Chang, H. 1986. Modelling the relationship of accidents to miles travelled.

Lee, J., and Mannering, F. 2002. Impact of Roadside Features

on The Frequency and Severity of Run-off-Roadway Accidents:

An Empirical Analysis. Accidents Analysis and Prevention, 34,

-161.

Miaou, S.P., and Lum, H. 1993. Modelling vehicle accidents and highway geometric design relationships. Accident Analysis and Prevention, 25 (http://dx.doi.org/10.1016/0001-4575(93)90034-t).

Omari-Sasu, A.Y, Adjei Mensah Isaac’, R.K Boadi. 2016. Statistical Models for Counting Data with Applicantions to Road Accidents in Ghana. International Journal of Statistical and Applications 6(3): 123-137.

Shankar, V., Mannering, F., and Barfield, W. 1995. Effect of Roadway Geometrics and Environmental Factors on Rural Freeway Accident Frequencies. Accidents Analysis and Prevention, 27, 371-389.




Flag Counter