Open AccessData analytics for traffic flow prediction in Custom using Long Short Term Memory (LSTM) networks
Author(s) -
Lee Pin Loon,
Elbaraa Refaie,
Ahmad Athif Mohd Faudzi
Publication year - 2021
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/2107/1/012006
Subject(s) - python (programming language) , computer science , timestamp , campus network , officer , traffic flow (computer networking) , term (time) , occupancy , analytics , data mining , real time computing , computer security , operating system , engineering , geography , architectural engineering , physics , archaeology , quantum mechanics
This paper proposes data analysis for traffic flow prediction of customs to help the officer in Customs, Immigration, and Quarantine (CIQ) Complex to understand more about the traffic situation in CIQ. Currently in CIQ, the traffic behaviour for car is unpredictable; sometimes the traffic is very heavy while there are times where all the lanes are cleared. There is a plan to have installation of cameras for smart traffic management system in the future. Therefore, this research aims to have prediction of traffic flow based on time and visualize the trend of traffic data for the officer. The data consist of traffic flow and the respective timestamp. To analyse it with time-series data, Long Short-Term Memory (LSTM) Recurrent Network is used as deep learning approach for prediction. The data pre-processing and training of model would be done using Python. To organize the data, Tableau Prep Builder is used and integrate with Python to publish the data to Tableau Server for storage. An interactive dashboard would be designed on Tableau and made available online for the usage of the officer.