In this Working Paper we analyse computational strategies for using aggregated spatio-temporal population data acquired from telecommunications networks to infer travel and movement patterns between geographical regions. Specifically, we focus on hour-by-hour cellphone counts for the SA-2 geographical regions covering the whole of New Zealand. This Working Paper describes the implementation of the inference algorithms, their ability to produce models of travel patterns during the day, and lays out opportunities for future development.
翻译:在本工作论文中,我们分析了使用从电信网络获取的聚合时空人口数据来推断地理区域之间的出行和运动模式的计算策略。具体而言,我们关注覆盖新西兰整个地区的SA-2地理区域的以小时为单位的手机计数。本工作论文描述了推断算法的实现,它们在白天产生出行模式的模型的能力,并概述了未来发展的机会。