Abstract
Due to the rapid development of communication and sensing technology, a large amount of mobile data is collected so that we can infer the complex movement laws of humans. For cities, some unusual events may endanger public safety. If the early warning of an abnormal event can be issued, it is of great application value to urban construction services. To detect urban anomalies, this paper proposes the Hierarchical Urban Anomaly Detection ( HUAD ) framework. The first step in this framework is to build rough anomaly characteristics that need to be calculated by some traffic flow consisted of subway and taxi data. In the second step, the alternative abnormal regions were obtained. Then, the long short-term memory (LSTM) network is used to predict the traffic to get the historical anomaly scores. Following that, the refined anomaly characteristics are generated from adjacent regions, adjacent periods and historical anomalies. The final abnormal regions were detected by One-Class Support Vector Machine (OC-SVM). At last, based on real data sets, we analyzes the traffic flow of the target region and adjacent regions from multiple perspectives in view of the large crowd gathering activities, and the effectiveness of the method is verified.
Highlights
With the development of communication technology and sensing technology, massive multi-source heterogeneous data are generated from clients, such as vehicle trajectories, social platform data, geographic information system (GIS) data [1], [2], etc
Gao et al proposed multimodal deep learning model based on spatio-temporal data witch can handle complex nonlinear urban traffic flow predictions with satisfactory accuracy and effectiveness [14]
ANOMALY DETECTION MODEL BASED ON One-Class Support Vector Machine (OC-SVM) The historical anomaly score are obtained by calculating the difference between the actual normal flow and the predicted flow from multi-step long short-term memory (LSTM)
Summary
With the development of communication technology and sensing technology, massive multi-source heterogeneous data are generated from clients, such as vehicle trajectories, social platform data, geographic information system (GIS) data [1], [2], etc. Anomalies in different regions and different time periods are constructed based on the similarity between different regions and different data sources [1], and the candidate anomaly regions are filtered out by the OC-SVM On this basis, to construct additional anomaly characteristics, we first adopt Long Short-Term Memory (LSTM) [8] to learn traffic data and predict the traffic flow in the last 4 time periods [9]. Gao et al proposed multimodal deep learning model based on spatio-temporal data witch can handle complex nonlinear urban traffic flow predictions with satisfactory accuracy and effectiveness [14]. In order to reveal the characteristics of regional traffic flow patterns in large road networks, He et al employ dictionary-based compression theory to identify the features of both spatial and temporal patterns by analyzing the multi-dimensional traffic-related data [21].
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
