Characterizing Regional Importance in Cities with Human Mobility Motifs in Metro Networks
Authors:
Shuyang Shi,
Ding Lyu,
Lin Wang,
Xiaofan Wang,
Guanrong Chen
Abstract:
Uncovering higher-order spatiotemporal dependencies within human mobility networks offers valuable insights into the analysis of urban structures. In most existing studies, human mobility networks are typically constructed by aggregating all trips without distinguishing who takes which specific trip. Instead, we claim individual mobility motifs, higher-order structures generated by daily trips of…
▽ More
Uncovering higher-order spatiotemporal dependencies within human mobility networks offers valuable insights into the analysis of urban structures. In most existing studies, human mobility networks are typically constructed by aggregating all trips without distinguishing who takes which specific trip. Instead, we claim individual mobility motifs, higher-order structures generated by daily trips of people, as fundamental units of human mobility networks. In this paper, we propose two network construction frameworks at the level of mobility motifs in characterizing regional importance in cities. Firstly, we enhance the structural dependencies within mobility motifs and proceed to construct mobility networks based on the enhanced mobility motifs. Secondly, taking inspiration from PageRank, we speculate that people would allocate values of importance to destinations according to their trip intentions. A motif-wise network construction framework is proposed based on the established mechanism. Leveraging large-scale metro data across cities, we construct three types of human mobility networks and characterize the regional importance by node importance indicators. Our comparison results suggest that the motif-based mobility network outperforms the classic mobility network, thus highlighting the efficacy of the introduced human mobility motifs. Finally, we demonstrate that the performance in characterizing the regional importance is significantly improved by our motif-wise framework.
△ Less
Submitted 7 May, 2024;
originally announced May 2024.
AutoRepair: Automated Repair for AI-Enabled Cyber-Physical Systems under Safety-Critical Conditions
Authors:
Deyun Lyu,
Jiayang Song,
Zhenya Zhang,
Zhijie Wang,
Tianyi Zhang,
Lei Ma,
Jianjun Zhao
Abstract:
Cyber-Physical Systems (CPS) have been widely deployed in safety-critical domains such as transportation, power and energy. Recently, there comes an increasing demand in employing deep neural networks (DNNs) in CPS for more intelligent control and decision making in sophisticated industrial safety-critical conditions, giving birth to the class of DNN controllers. However, due to the inherent uncer…
▽ More
Cyber-Physical Systems (CPS) have been widely deployed in safety-critical domains such as transportation, power and energy. Recently, there comes an increasing demand in employing deep neural networks (DNNs) in CPS for more intelligent control and decision making in sophisticated industrial safety-critical conditions, giving birth to the class of DNN controllers. However, due to the inherent uncertainty and opaqueness of DNNs, concerns about the safety of DNN-enabled CPS are also surging. In this work, we propose an automated framework named AutoRepair that, given a safety requirement, identifies unsafe control behavior in a DNN controller and repairs them through an optimization-based method. Having an unsafe signal of system execution, AutoRepair iteratively explores the control decision space and searches for the optimal corrections for the DNN controller in order to satisfy the safety requirements. We conduct a comprehensive evaluation of AutoRepair on 6 instances of industry-level DNN-enabled CPS from different safety-critical domains. Evaluation results show that AutoRepair successfully repairs critical safety issues in the DNN controllers, and significantly improves the reliability of CPS.
△ Less
Submitted 12 April, 2023;
originally announced April 2023.
