The smart parking system is one of the most important problems in smart cities. Recently, an increasing number of sensors installed in parking spaces provide big spatio-temporal data which be used to analyse the parking situations in the city and help parking officers monitor the parking violations. Travelling Officer Problem was customised to formulate a path-finding problem that aims to maximise the probability of catching overstayed cars before they leave. One of the challenges is to extract effective features from the big spatio-temporal data and provide a data-driven solution to replace conventional solutions such as a simple rule-based system or single optimisation methods. In this paper, we propose a seamless end-to-end learning and optimisation framework that combines Long Short-Term Memory (LSTM) Auto-Encoder neural network, clustering and path-finding methods to solve the Travelling Officer Problem. Our extensive comparison experiments on a large scale real-world dataset have shown that our proposed solution outperforms any other single step methods or optimisation methods.
@article{DBLP:journals/tsas/ShaoTZQHCS20,
author = {Wei Shao and
Siyu Tan and
Sichen Zhao and
Kyle Kai Qin and
Xinhong Hei and
Jeffrey Chan and
Flora D. Salim},
bibsource = {dblp computer science bibliography, https://dblp.org},
biburl = {https://dblp.org/rec/journals/tsas/ShaoTZQHCS20.bib},
doi = {10.1145/3380966},
journal = {ACM Trans. Spatial Algorithms and Systems},
number = {3},
pages = {20:1--20:21},
timestamp = {Fri, 05 Jun 2020 01:00:00 +0200},
title = {Incorporating LSTM Auto-Encoders in Optimizations to Solve Parking
Officer Patrolling Problem},
url = {https://doi.org/10.1145/3380966},
volume = {6},
year = {2020}
}
© 2021 Flora Salim - CRUISE Research Group.