Bayesian flexible modeling of trip durations

Hugh Chipman, Edward George, Jason Lemp, Robert McCulloch

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

Recent advances in Bayesian modeling have led to stunning improvements in our ability to flexibly and easily model complex high-dimensional data. Flexibility comes from the use of a very large number of parameters without fixed dimension. Priors are placed on the parameters to avoid over-fitting and sensibly guide the search in model space for appropriate data-driven model choice. Modern computational, high dimensional search methods (in particular Markov Chain Monte Carlo) then allow us to search the parameter space. This paper introduces the application of BART, Bayesian Additive Regression Trees, to modelling trip durations. We have survey data on characteristics of trips in the Austin area. We seek to relate the trip duration to features of the household and trip characteristics. BART enables one to make inferences about the relationship with minimal assumptions and user decisions.

Original languageEnglish (US)
Pages (from-to)686-698
Number of pages13
JournalTransportation Research Part B: Methodological
Volume44
Issue number5
DOIs
StatePublished - Jun 2010
Externally publishedYes

Fingerprint

Markov processes
flexibility
regression
ability
Modeling
Household
Survey data
Model choice
Markov chain Monte Carlo
Bayesian modeling
Overfitting
Regression tree
Inference

Keywords

  • Boosting
  • Ensemble modeling
  • Markov Chain Monte Carlo

ASJC Scopus subject areas

  • Management Science and Operations Research
  • Transportation

Cite this

Bayesian flexible modeling of trip durations. / Chipman, Hugh; George, Edward; Lemp, Jason; McCulloch, Robert.

In: Transportation Research Part B: Methodological, Vol. 44, No. 5, 06.2010, p. 686-698.

Research output: Contribution to journalArticle

Chipman, Hugh ; George, Edward ; Lemp, Jason ; McCulloch, Robert. / Bayesian flexible modeling of trip durations. In: Transportation Research Part B: Methodological. 2010 ; Vol. 44, No. 5. pp. 686-698.
@article{e93d431d526140b4b21d486469dc2289,
title = "Bayesian flexible modeling of trip durations",
abstract = "Recent advances in Bayesian modeling have led to stunning improvements in our ability to flexibly and easily model complex high-dimensional data. Flexibility comes from the use of a very large number of parameters without fixed dimension. Priors are placed on the parameters to avoid over-fitting and sensibly guide the search in model space for appropriate data-driven model choice. Modern computational, high dimensional search methods (in particular Markov Chain Monte Carlo) then allow us to search the parameter space. This paper introduces the application of BART, Bayesian Additive Regression Trees, to modelling trip durations. We have survey data on characteristics of trips in the Austin area. We seek to relate the trip duration to features of the household and trip characteristics. BART enables one to make inferences about the relationship with minimal assumptions and user decisions.",
keywords = "Boosting, Ensemble modeling, Markov Chain Monte Carlo",
author = "Hugh Chipman and Edward George and Jason Lemp and Robert McCulloch",
year = "2010",
month = "6",
doi = "10.1016/j.trb.2010.01.007",
language = "English (US)",
volume = "44",
pages = "686--698",
journal = "Transportation Research, Series B: Methodological",
issn = "0191-2615",
publisher = "Elsevier Limited",
number = "5",

}

TY - JOUR

T1 - Bayesian flexible modeling of trip durations

AU - Chipman, Hugh

AU - George, Edward

AU - Lemp, Jason

AU - McCulloch, Robert

PY - 2010/6

Y1 - 2010/6

N2 - Recent advances in Bayesian modeling have led to stunning improvements in our ability to flexibly and easily model complex high-dimensional data. Flexibility comes from the use of a very large number of parameters without fixed dimension. Priors are placed on the parameters to avoid over-fitting and sensibly guide the search in model space for appropriate data-driven model choice. Modern computational, high dimensional search methods (in particular Markov Chain Monte Carlo) then allow us to search the parameter space. This paper introduces the application of BART, Bayesian Additive Regression Trees, to modelling trip durations. We have survey data on characteristics of trips in the Austin area. We seek to relate the trip duration to features of the household and trip characteristics. BART enables one to make inferences about the relationship with minimal assumptions and user decisions.

AB - Recent advances in Bayesian modeling have led to stunning improvements in our ability to flexibly and easily model complex high-dimensional data. Flexibility comes from the use of a very large number of parameters without fixed dimension. Priors are placed on the parameters to avoid over-fitting and sensibly guide the search in model space for appropriate data-driven model choice. Modern computational, high dimensional search methods (in particular Markov Chain Monte Carlo) then allow us to search the parameter space. This paper introduces the application of BART, Bayesian Additive Regression Trees, to modelling trip durations. We have survey data on characteristics of trips in the Austin area. We seek to relate the trip duration to features of the household and trip characteristics. BART enables one to make inferences about the relationship with minimal assumptions and user decisions.

KW - Boosting

KW - Ensemble modeling

KW - Markov Chain Monte Carlo

UR - http://www.scopus.com/inward/record.url?scp=77952954024&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77952954024&partnerID=8YFLogxK

U2 - 10.1016/j.trb.2010.01.007

DO - 10.1016/j.trb.2010.01.007

M3 - Article

AN - SCOPUS:77952954024

VL - 44

SP - 686

EP - 698

JO - Transportation Research, Series B: Methodological

JF - Transportation Research, Series B: Methodological

SN - 0191-2615

IS - 5

ER -