@article{ref1,
title="Multi sensor, multivariate, and multi-class incident detection system for arterial streets",
journal="Transportation and traffic theory",
year="1996",
author="Thomas, Nigel",
volume="13",
number="",
pages="315-339",
abstract="A novel approach to incident detection on arterial streets that utilizes multi-sensor, multi-class, and multivariate classifiers to differentiate between various traffic states is proposed. The similarities between the Bayes' fusion of multi-sensor allocations and Multiple Attribute Decision Making (MADM) are established. An array of MADM algorithms is thus made available to the traffic engineer for purposes of fusion of multi-sensor allocations. One such algorithm is applied to the detection of incidents on arterial streets using detector occupancies and vehicle counts by lane, probe travel times, and probe report numbers as attributes. The probe data proves valuable in enhancing the performance of detector data based models. Models based solely on probe data lack in performance, due to excessive overlaps in class distributions. The possibilities for identifying incidents through their flow imbalance impacts, using multivariate detector classifiers, prove promising. (A) For the covering abstract see IRRD 886400.<p />",
language="en",
issn="",
doi="",
url="http://dx.doi.org/"
}