Efficient Transport Simulation With Restricted Batch-Mode Active Learning

Francisco Antunes*, Bernardete Ribeiro, Francisco C. Pereira, Rui Gomes

*Corresponding author for this work

Research output: Contribution to journalJournal articleResearchpeer-review

Abstract

Simulation modeling is a well-known and recurrent approach to study the performance of urban systems. Taking into account the recent and continuous transformations within increasingly complex and multidimensional cities, the use of simulation tools is, in many cases, the only feasible and reliable approach to analyze such dynamic systems. However, simulation models can become very time consuming when detailed input-space exploration is needed. To tackle this problem, simulation metamodels are often used to approximate the simulators' results. In this paper, we propose an active learning algorithm based on the Gaussian process (GP) framework that gathers the most informative simulation data points in batches, according to both their predictive variances and to the relative distance between them. This allows us to explore the simulators' input space with fewer data points and in parallel, and thus in a more efficient way, while avoiding computationally expensive simulation runs in the process. We take advantage of the closeness notion encoded into the GP to select batches of points in such a way that they do not belong to the same high-variance neighborhoods. In addition, we also suggest two simple and practical user-defined stopping criteria so that the iterative learning procedure can be fully automated. We illustrate this methodology using three experimental settings. The results show that the proposed methodology is able to improve the exploration efficiency of the simulation input space in comparison with non-restricted batch-mode active learning procedures.

Original languageEnglish
JournalI E E E Transactions on Intelligent Transportation Systems
Volume19
Issue number11
Pages (from-to)3642-3651
Number of pages10
ISSN1524-9050
DOIs
Publication statusPublished - 2018

Keywords

  • Active learning
  • Gaussian processes
  • Simulation metamodels
  • Transport simulation

Cite this

@article{39f6602bc059498ca465b668ba8bd828,
title = "Efficient Transport Simulation With Restricted Batch-Mode Active Learning",
abstract = "Simulation modeling is a well-known and recurrent approach to study the performance of urban systems. Taking into account the recent and continuous transformations within increasingly complex and multidimensional cities, the use of simulation tools is, in many cases, the only feasible and reliable approach to analyze such dynamic systems. However, simulation models can become very time consuming when detailed input-space exploration is needed. To tackle this problem, simulation metamodels are often used to approximate the simulators' results. In this paper, we propose an active learning algorithm based on the Gaussian process (GP) framework that gathers the most informative simulation data points in batches, according to both their predictive variances and to the relative distance between them. This allows us to explore the simulators' input space with fewer data points and in parallel, and thus in a more efficient way, while avoiding computationally expensive simulation runs in the process. We take advantage of the closeness notion encoded into the GP to select batches of points in such a way that they do not belong to the same high-variance neighborhoods. In addition, we also suggest two simple and practical user-defined stopping criteria so that the iterative learning procedure can be fully automated. We illustrate this methodology using three experimental settings. The results show that the proposed methodology is able to improve the exploration efficiency of the simulation input space in comparison with non-restricted batch-mode active learning procedures.",
keywords = "Active learning, Gaussian processes, Simulation metamodels, Transport simulation",
author = "Francisco Antunes and Bernardete Ribeiro and Pereira, {Francisco C.} and Rui Gomes",
year = "2018",
doi = "10.1109/TITS.2018.2842695",
language = "English",
volume = "19",
pages = "3642--3651",
journal = "I E E E Transactions on Intelligent Transportation Systems",
issn = "1524-9050",
publisher = "Institute of Electrical and Electronics Engineers",
number = "11",

}

Efficient Transport Simulation With Restricted Batch-Mode Active Learning. / Antunes, Francisco; Ribeiro, Bernardete; Pereira, Francisco C.; Gomes, Rui.

In: I E E E Transactions on Intelligent Transportation Systems, Vol. 19, No. 11, 2018, p. 3642-3651.

Research output: Contribution to journalJournal articleResearchpeer-review

TY - JOUR

T1 - Efficient Transport Simulation With Restricted Batch-Mode Active Learning

AU - Antunes, Francisco

AU - Ribeiro, Bernardete

AU - Pereira, Francisco C.

AU - Gomes, Rui

PY - 2018

Y1 - 2018

N2 - Simulation modeling is a well-known and recurrent approach to study the performance of urban systems. Taking into account the recent and continuous transformations within increasingly complex and multidimensional cities, the use of simulation tools is, in many cases, the only feasible and reliable approach to analyze such dynamic systems. However, simulation models can become very time consuming when detailed input-space exploration is needed. To tackle this problem, simulation metamodels are often used to approximate the simulators' results. In this paper, we propose an active learning algorithm based on the Gaussian process (GP) framework that gathers the most informative simulation data points in batches, according to both their predictive variances and to the relative distance between them. This allows us to explore the simulators' input space with fewer data points and in parallel, and thus in a more efficient way, while avoiding computationally expensive simulation runs in the process. We take advantage of the closeness notion encoded into the GP to select batches of points in such a way that they do not belong to the same high-variance neighborhoods. In addition, we also suggest two simple and practical user-defined stopping criteria so that the iterative learning procedure can be fully automated. We illustrate this methodology using three experimental settings. The results show that the proposed methodology is able to improve the exploration efficiency of the simulation input space in comparison with non-restricted batch-mode active learning procedures.

AB - Simulation modeling is a well-known and recurrent approach to study the performance of urban systems. Taking into account the recent and continuous transformations within increasingly complex and multidimensional cities, the use of simulation tools is, in many cases, the only feasible and reliable approach to analyze such dynamic systems. However, simulation models can become very time consuming when detailed input-space exploration is needed. To tackle this problem, simulation metamodels are often used to approximate the simulators' results. In this paper, we propose an active learning algorithm based on the Gaussian process (GP) framework that gathers the most informative simulation data points in batches, according to both their predictive variances and to the relative distance between them. This allows us to explore the simulators' input space with fewer data points and in parallel, and thus in a more efficient way, while avoiding computationally expensive simulation runs in the process. We take advantage of the closeness notion encoded into the GP to select batches of points in such a way that they do not belong to the same high-variance neighborhoods. In addition, we also suggest two simple and practical user-defined stopping criteria so that the iterative learning procedure can be fully automated. We illustrate this methodology using three experimental settings. The results show that the proposed methodology is able to improve the exploration efficiency of the simulation input space in comparison with non-restricted batch-mode active learning procedures.

KW - Active learning

KW - Gaussian processes

KW - Simulation metamodels

KW - Transport simulation

U2 - 10.1109/TITS.2018.2842695

DO - 10.1109/TITS.2018.2842695

M3 - Journal article

AN - SCOPUS:85050638079

VL - 19

SP - 3642

EP - 3651

JO - I E E E Transactions on Intelligent Transportation Systems

JF - I E E E Transactions on Intelligent Transportation Systems

SN - 1524-9050

IS - 11

ER -