The capacity to precisely figure the development of road traffic is vital in traffic management and control applications. Urban traffic gridlock is a standout amongst the most extreme issues of regular day to day existence in Metropolitan regions. To manage this issue, variations in rush hour patterns can be exceptionally valuable in helping Intelligent Transport System (ITS) which furnishes drivers with helpful and precise route and travel-time data. In this research work, data storage and analysis is performed in Hive data warehouse. HiveQL Language is used for analyzing the features of traffic data and these analyzed data are processed in SPARK framework. Here with the help of Mlib library, Decision tree classifier algorithm is used for classifying the traffic to low, medium and high traffic range. Using the classifier, abnormality in traffic is detected and it is notified to the user through an android application. Finally the best route with high accuracy is predicted which can be accessed by the drivers. More specifically they can help the user’s to choose a much better and less time-consuming path than their original intended route. Experimental result compares forecast exactness of the proposed model with the existing approaches and achieves 95% of accuracy. From the experiments, the results illustrate the proposed approach is unique and robust compared to other models.
Traffic Prediction Using Decision Tree Classifier in Hive Metastore
Lect. Notes on Data Eng. and Comms.Technol.
International conference on Computer Networks, Big data and IoT ; 2018 ; Madurai, India December 19, 2018 - December 20, 2018
Proceeding of the International Conference on Computer Networks, Big Data and IoT (ICCBI - 2018) ; Chapter : 68 ; 571-578
2019-08-01
8 pages
Article/Chapter (Book)
Electronic Resource
English
TIBKAT | 1.1871,6; mehr nicht digitalisiert
Traffic Incident Duration Prediction Based on the Bayesian Decision Tree Method
British Library Conference Proceedings | 2008
|