CGAN-EB: A Non-parametric Empirical Bayes Method for Crash Hotspot Identification Using Conditional Generative Adversarial Networks: A Real-world Crash Data Study

by   Mohammad Zarei, et al.

The empirical Bayes (EB) method based on parametric statistical models such as the negative binomial (NB) has been widely used for ranking sites in road network safety screening process. This paper is the continuation of the authors previous research, where a novel non-parametric EB method for modelling crash frequency data data based on Conditional Generative Adversarial Networks (CGAN) was proposed and evaluated over several simulated crash data sets. Unlike parametric approaches, there is no need for a pre-specified underlying relationship between dependent and independent variables in the proposed CGAN-EB and they are able to model any types of distributions. The proposed methodology is now applied to a real-world data set collected for road segments from 2012 to 2017 in Washington State. The performance of CGAN-EB in terms of model fit, predictive performance and network screening outcomes is compared with the conventional approach (NB-EB) as a benchmark. The results indicate that the proposed CGAN-EB approach outperforms NB-EB in terms of prediction power and hotspot identification tests.


Crash Data Augmentation Using Conditional Generative Adversarial Networks (CGAN) for Improving Safety Performance Functions

In this paper, we present a crash frequency data augmentation method bas...

Screening Methods for Classification Based on Non-parametric Bayesian Tests

Feature or variable selection is a problem inherent to large data sets. ...

Joint Estimation of the Non-parametric Transitivity and Preferential Attachment Functions in Scientific Co-authorship Networks

We propose a statistical method to estimate simultaneously the non-param...

Spatial multiresolution analysis approach to identify crash hotspots and estimate crash risk

In this paper, the authors evaluate the performance of a spatial multire...

Explaining predictive models using Shapley values and non-parametric vine copulas

The original development of Shapley values for prediction explanation re...

Likelihood-based Changepoint Detection in Preferential Attachment Networks

Generative, temporal network models play an important role in analyzing ...

Please sign up or login with your details

Forgot password? Click here to reset