Comparison of neural networks and gravity models in trip distribution

Research output: Contribution to journalArticleAcademicpeer-review

18 Citations (Scopus)
36 Downloads (Pure)

Abstract

Transportation engineers are commonly faced with the question of how to extract information from expensive and scarce field data. Modeling the distribution of trips between zones is complex and dependent on the quality and availability of field data. This research explores the performance of neural networks in trip distribution modeling and compares the results with commonly used doubly constrained gravity models. The approach differs from other research in several respects; the study is based on both synthetic data, varying in complexity, as well as real-world data. Furthermore, neural networks and gravity models are calibrated using different percentages of hold out data. Extensive statistical analyses are conducted to obtain necessary sample sizes for significant results. The results show that neural networks outperform gravity models when data are scarce in both synthesized as well as real-world cases. Sample size for statistically significant results is forty times lower for neural networks.
Original languageEnglish
Pages (from-to)104-119
Number of pages15
JournalComputer-aided civil and infrastructure engineering
Volume21
Issue number2
DOIs
Publication statusPublished - 2006

Fingerprint

Gravitation
gravity
Neural networks
Availability
Engineers
modeling
distribution
comparison

Keywords

  • METIS-213425
  • IR-58422

Cite this

@article{fcbb7896b57d451a9a4060c6c1954014,
title = "Comparison of neural networks and gravity models in trip distribution",
abstract = "Transportation engineers are commonly faced with the question of how to extract information from expensive and scarce field data. Modeling the distribution of trips between zones is complex and dependent on the quality and availability of field data. This research explores the performance of neural networks in trip distribution modeling and compares the results with commonly used doubly constrained gravity models. The approach differs from other research in several respects; the study is based on both synthetic data, varying in complexity, as well as real-world data. Furthermore, neural networks and gravity models are calibrated using different percentages of hold out data. Extensive statistical analyses are conducted to obtain necessary sample sizes for significant results. The results show that neural networks outperform gravity models when data are scarce in both synthesized as well as real-world cases. Sample size for statistically significant results is forty times lower for neural networks.",
keywords = "METIS-213425, IR-58422",
author = "Frans Tillema and {van Zuilekom}, {Kasper M.} and {van Maarseveen}, M.F.A.M.",
year = "2006",
doi = "10.1111/j.1467-8667.2005.00421.x",
language = "English",
volume = "21",
pages = "104--119",
journal = "Computer-aided civil and infrastructure engineering",
issn = "1093-9687",
publisher = "Wiley-Blackwell",
number = "2",

}

Comparison of neural networks and gravity models in trip distribution. / Tillema, Frans; van Zuilekom, Kasper M.; van Maarseveen, M.F.A.M.

In: Computer-aided civil and infrastructure engineering, Vol. 21, No. 2, 2006, p. 104-119.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - Comparison of neural networks and gravity models in trip distribution

AU - Tillema, Frans

AU - van Zuilekom, Kasper M.

AU - van Maarseveen, M.F.A.M.

PY - 2006

Y1 - 2006

N2 - Transportation engineers are commonly faced with the question of how to extract information from expensive and scarce field data. Modeling the distribution of trips between zones is complex and dependent on the quality and availability of field data. This research explores the performance of neural networks in trip distribution modeling and compares the results with commonly used doubly constrained gravity models. The approach differs from other research in several respects; the study is based on both synthetic data, varying in complexity, as well as real-world data. Furthermore, neural networks and gravity models are calibrated using different percentages of hold out data. Extensive statistical analyses are conducted to obtain necessary sample sizes for significant results. The results show that neural networks outperform gravity models when data are scarce in both synthesized as well as real-world cases. Sample size for statistically significant results is forty times lower for neural networks.

AB - Transportation engineers are commonly faced with the question of how to extract information from expensive and scarce field data. Modeling the distribution of trips between zones is complex and dependent on the quality and availability of field data. This research explores the performance of neural networks in trip distribution modeling and compares the results with commonly used doubly constrained gravity models. The approach differs from other research in several respects; the study is based on both synthetic data, varying in complexity, as well as real-world data. Furthermore, neural networks and gravity models are calibrated using different percentages of hold out data. Extensive statistical analyses are conducted to obtain necessary sample sizes for significant results. The results show that neural networks outperform gravity models when data are scarce in both synthesized as well as real-world cases. Sample size for statistically significant results is forty times lower for neural networks.

KW - METIS-213425

KW - IR-58422

U2 - 10.1111/j.1467-8667.2005.00421.x

DO - 10.1111/j.1467-8667.2005.00421.x

M3 - Article

VL - 21

SP - 104

EP - 119

JO - Computer-aided civil and infrastructure engineering

JF - Computer-aided civil and infrastructure engineering

SN - 1093-9687

IS - 2

ER -