Event Detection In Communication And Transportation Data
PATTERN RECOGNITION AND IMAGE ANALYSIS, IBPRIA 2013(2013)
摘要
The digitalization of the urban infrastructure and pervasiveness of mobile handsets allows to model the movement and communication patterns of people in the city. In this paper, we focus on automatically detecting unusual behavior (events) in the city of Barcelona by analyzing aggregated data from two urban sensors: cell phone towers and shared bicycle rentals. We use a state-of-the-art novelty detection algorithm and compare its performance to two alternative approaches. In addition, we design a simple tree classifier to distinguish between three types of unusual days: rainy days, holidays and days with special events.
更多查看译文
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络