Abstract
Road sensors gather a lot of statistical data about traffic. In this
paper, we discuss how a measure for the amount of traffic on the roads
can be derived from this data, such that the measure is independent
of the number and placement of sensors, and the calculations can be
performed quickly for large amounts of data.
We discuss how a graph of the road sensors can be constructed, and
how the number of cars and car-kilometers can be estimated on this
graph. Further, methods for dealing with missing data are presented,
and the benefits of principal component analysis are discussed.
paper, we discuss how a measure for the amount of traffic on the roads
can be derived from this data, such that the measure is independent
of the number and placement of sensors, and the calculations can be
performed quickly for large amounts of data.
We discuss how a graph of the road sensors can be constructed, and
how the number of cars and car-kilometers can be estimated on this
graph. Further, methods for dealing with missing data are presented,
and the benefits of principal component analysis are discussed.
| Original language | English |
|---|---|
| Title of host publication | Proceedings 98th European Study Group Mathematics with Industry Delft 2014 |
| Publisher | Technische Universiteit Delft |
| Pages | 2-22 |
| Number of pages | 21 |
| ISBN (Print) | 978-94-6186-306-5 |
| Publication status | Published - 2014 |
Keywords
- traffic index
- road sensor
- graph construction
- statistical imputation
- principal component analysis
- CUR decomposition