Pemodelan Data Kecelakaan pada Perlintasan Sebidang Kereta Api DAOP VII Madiun

Abstract

The study aims to find the most suitable regression model of railroad level crossings accident data to obtain factors significantly affect the number of fatalities in DAOP VII Madiun. The variables used were road width, right slope, left slope, train frequency, type of crossing, large kilometer angle of view, small kilometer angle of view, Early Warning System, road status, type of crossing gates, left and right caution sign, and rumble band markings. The study used comparing method between the results of the Poisson regression model and the Negative Binomial regression model. It is found that the Negative Binomial regression model has a smaller Akaike Information Criteria and Root Mean Square Error than the Poisson model. It can be concluded that the Negative Binomial regression model is a better choice in modeling accident data in DAOP VII Madiun level crossing. Based on the modeling result, the factors significantly affect the number of fatalities in level crossings accident are road width, road status, type of crossing gate, and the presence or absence of caution signs.