Georgia Institute of Technology
TRAFFIC CONGESTION MODELING WITH DEEP ATTENTION HAWKES PROCESS
Abstract
dc:description.abstractIn this thesis, we focus on modeling the traffic congestion in the city of Atlanta. We are trying to predict future congestion events on the main highways in Atlanta. We present a novel framework for modeling traffic congestion events over road networks based on mutually exciting Spatio-temporal point process models. We use multi-modal data by combining traffic sensor networks data with police reports, which contain two types of triggering mechanisms for congestion events. To capture the non-homogeneous temporal dependence of the event on the past, we introduce a novel attention-based approach for the point process model. To incorporate the directional spatial dependence induced by the road network, we adapt the "tail-up" model from the spatial statistics context. We demonstrate the superior performance of our approach compared to the state-of-the-art for both synthetic and real data.
Degree
thesis:*- Level thesis:degree_level
- Masters
- Department dc:contributor.department
- Electrical and Computer Engineering
- Grantor dc:publisher
- Georgia Institute of Technology
- Year dc:date.issued
- 2020
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Ding, Ruyi
- Advisor dc:contributor.advisor
-
- Xie, Yao
- Committee members dc:contributor.committeemember
-
- Davenport, Mark
- Bloch, Matthieu
Subjects
dc:subject × 3Identifiers
dc:identifier.*- Handle dc:identifier.uri
- http://hdl.handle.net/1853/64648