Outlier Detection in Urban Traffic Data

摘要: This paper provides a summary of the tutorial on outlier detection in urban traffic data. We present existing solutions three main categories: statistical techniques, similarity-based and techniques based pattern analysis. The first category groups employing models to identify anomalies second using distance measures neighborhoods derive local density estimates. third explores correlation between flow values by concepts from explain discuss example for each category, we outline perspectives open questions research challenges. relate general view notion locality, i.e., context reference used definition comparison outlierness, order gain better understanding intuition, limitations, benefits various methods traffic. way, hope provide some guidance practitioners selecting most suitable their case.