Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
6937344 | Transportation Research Part C: Emerging Technologies | 2012 | 16 Pages |
Abstract
⺠Traffic flows consist of: trend, Gaussian type fluctuations and “bursts”. ⺠We improve predictions by retrieving intra-day trend prior to prediction. ⺠We cannot predict the “bursts”, due to smoothness assumption of prediction models. ⺠We can use PPCA based imputation method to mitigate data missing problem. ⺠We cannot predict but can identify traffic breakdowns.
Related Topics
Physical Sciences and Engineering
Computer Science
Computer Science Applications
Authors
Chenyi Chen, Yin Wang, Li Li, Jianming Hu, Zuo Zhang,