Predicting traffic incident duration using deep learning model with real-time data
Traffic accidents have a negative impact on traffic. The prediction of incident clearance time helps to reduce its impact by diverting traffic flow, assisting traffic management organizations in making decisions about appropriate responses and resource allocation and identifying critical factors tha...
Main Author: | |
---|---|
Other Authors: | |
Format: | Final Year Project (FYP) |
Language: | English |
Published: |
2019
|
Subjects: | |
Online Access: | http://hdl.handle.net/10356/77626 |
Summary: | Traffic accidents have a negative impact on traffic. The prediction of incident clearance time helps to reduce its impact by diverting traffic flow, assisting traffic management organizations in making decisions about appropriate responses and resource allocation and identifying critical factors that influence the length of duration. Previous researches utilized more detailed information that may be confidential to the public. Therefore, this study aimed to predict duration with transparent real-time data from LTA Datamall, MSS and OpenStreetMap. This study used a deep learning model to predict the incident duration. Various tests were carried out to optimize the neural network and to achieve the possible highest accuracy. The result of this research was comparable to previous researches in terms of MAE and MAPE, improvement in accuracy was be observed. This research also pointed out the directions for future research. |
---|