Predicting driving direction with weighted markov model
conference contribution
posted on 2012-12-01, 00:00 authored by B Mao, J Cao, Z Wu, Guangyan HuangGuangyan Huang, J LiDriving direction prediction can be useful in different applications such as driver warning and route recommendation. In this paper, a framework is proposed to predict the driving direction based on weighted Markov model. First the city POI (Point of Interesting) map is generated from trajectory data using weighted PageRank algorithm. Then, a weighted Markov model is trained for the near term driving direction prediction based on the POI map and historical trajectories. The experimental results on real-world data set indicate that the proposed method can improve the original Markov prediction model by 10% at some circumstances and 5% overall. © Springer-Verlag 2012.
History
Volume
7713Pagination
407-418Location
Nanjing, ChinaPublisher DOI
Start date
2012-12-15End date
2012-12-18ISSN
0302-9743eISSN
1611-3349ISBN-13
9783642355271Language
EngPublication classification
E Conference publication, E1.1 Full written paper - refereedCopyright notice
2013, SpringerEditor/Contributor(s)
Zhou S, Zhang S, Karypis GTitle of proceedings
Advanced Data Mining and ApplicationsEvent
International Conference on Advanced Data Mining and Applications (8th : 2012 : Nanjing, China)Publisher
Springer-VerlagPlace of publication
Berlin, GermanySeries
Lecture Notes in Artificial IntelligenceUsage metrics
Categories
No categories selectedKeywords
Licence
Exports
RefWorksRefWorks
BibTeXBibTeX
Ref. managerRef. manager
EndnoteEndnote
DataCiteDataCite
NLMNLM
DCDC