Work

Modeling Cycling Counts Using Crowd-Sourced Data

Year: 2018
Type: dissertation
Abstract: Cycling is beneficial for an individual’s physical and mental well being. Additionally, cycling is a means of sustainable transportation. A critical issue in alternative transportation studies is understanding how new data sources detailing cycling volumes can be used to augment traditional cycling data such as surveys and manual counts. This research addresses the question: “How can cycling flows in specific city corridors be understood and modeled through the fusion of crowdsourced data and traditional data sources?” The results of this work synthesize crowd-sourced and traditional data sources and create a modeling framework to predict the volume of cyclists at a city corridor level. The proposed modeling framework uses a Long Short Term Memory Model, a type Artificial Neural Network, that is able to handle large volumes of spatial data, and account for spatial and temporal autocorrelation.
Language: English
Subfield: Transportation
Open Access status: green
OpenAlex ID: W2954170632
Cites: 74
Cited by:
Related to: 10