Machine Learning Algorithms for Traffic Interruption Detection

Abstract

Detection of traffic interruptions is a critical aspect of managing traffic on urban road networks. This work outlines a semi-supervised strategy to automatically detect traffic interruptions occurring on arteries using high resolution data from widely deployed inductive loop detectors. The techniques highlighted in this paper are tested on data collected from detectors installed on more than 300 signalized intersections over a 6 month period. Our results show that we can detect interruptions with high precision and recall.