Multilevel Dirichlet process mixture analysis of railway grade crossing crash data
详细信息 查看全文
- 作者：Shahram Heydaria ; shahram.heydari@uwaterloo.ca" class="auth_mail" title="E-mail the corresponding author ; Liping Fua ; 1 ; lfu@uwaterloo.ca" class="auth_mail" title="E-mail the corresponding author ; Dominique Lordb ; 2 ; d-lord@tamu.edu" class="auth_mail" title="E-mail the corresponding author ; Bani K. Mallickc ; 3 ; bmallick@stat.tamu.edu" class="auth_mail" title="E-mail the corresponding author
- 关键词：Unobserved heterogeneity ; Latent subpopulations ; Random effects models ; Finite mixture models ; Spatial/regional multilevel models ; Dirichlet process mixture models
- 刊名：Analytic Methods in Accident Research
- 出版年：2016
- 出版时间：March 2016
- 年：2016
- 卷：9
- 期：Complete
- 页码：27-43
- 全文大小：870 K
文摘
- class="listitem" id="list_li0005">
- class="label">•
We employ a flexible multilevel model using a Dirichlet process mixture.
- class="label">•
The standard parametric assumption in the random effects model is relaxed.
- class="label">•
The expected number of latent components in data is estimated as part of the analysis.
- class="label">•
Latent subpopulations and outliers are identified among different Canadian regions.