Urban Anomalies
收藏资源简介:
Urban Anomalies是一个模拟人类移动行为的数据集,由埃默里大学创建。该数据集包含1000个代理的模拟数据,涵盖了正常和异常行为阶段。数据集通过改变代理的行为逻辑,注入了四种异常行为类型:饥饿、工作、社交和兴趣异常。创建过程中采用了三种异常注入方法:集中式、传染病模型和基于位置的模型。数据集旨在解决基于位置的异常检测问题,适用于公共健康、安全、福利和城市规划等领域。
Urban Anomalies is a dataset for simulating human mobility behavior, developed by Emory University. It contains simulated data from 1000 agents, covering both normal and anomalous behavioral phases. Four types of anomalous behaviors—hunger, work, social, and interest anomalies—are injected by modifying the agents' underlying behavioral logic. Three anomaly injection approaches are employed during the dataset construction: centralized, epidemic model, and location-based model. This dataset targets location-based anomaly detection tasks, and finds applications in fields including public health, safety, social welfare, and urban planning.

- 1Urban Anomalies: A Simulated Human Mobility Dataset with Injected Anomalies埃默里大学 · 2024年



