A new spatial count data model with Bayesian additive regression trees for accident hot spot identification. (September 2020)

• Record Type: Journal Article
• Authors: Krueger, Rico; Bansal, Prateek; Buddhavarapu, Prasad
• Is Part Of: Accident analysis and prevention. Volume 144(2020)
• Publication Date: 2020-09
• Subjects: Accident analysis -- Site ranking -- Spatial count data modelling -- Negative binomial model -- Bayesian additive regression trees -- Pólya-Gamma data augmentation; Accidents -- Prevention -- Periodicals; Accident Prevention -- Periodicals; Accidents -- Prévention -- Périodiques; 363.106
• DOI: 10.1016/j.aap.2020.105623
• Languages: English
• ISSNs: 0001-4575

Abstract

Highlights: New spatial count data model with flexible link function specification. Additive regression trees enable endogenous partitioning of predictor space. MCMC algorithm for fully Bayesian inference. New method offers excellent goodness of fit and site ranking ability. Abstract: The identification of accident hot spots is a central task of road safety management. Bayesian count data models have emerged as the workhorse method for producing probabilistic rankings of hazardous sites in road networks. Typically, these methods assume simple linear link function specifications, which, however, limit the predictive power of a model. Furthermore, extensive specification searches are precluded by complex model structures arising from the need to account for unobserved heterogeneity and spatial correlations. Modern machine learning (ML) methods offer ways to automate the specification of the link function. However, these methods do not capture estimation uncertainty, and it is also difficult to incorporate spatial correlations. In light of these gaps in the literature, this paper proposes a new spatial negative binomial model which uses Bayesian additive regression trees to endogenously select the specification of the link function. Posterior inference in the proposed model is made feasible with the help of the Pólya-Gamma data augmentation technique. We test the performance of this new model on a crash count data set from a metropolitan highway network. The empirical results show that the proposed model performs at least as well as a baseline spatial count data model with random parameters in terms of goodness of fit and site ranking ability.