University of Twente Student Theses


Forecasting bus ridership with trip planner usage data : a machine learning application

Roosmalen, J.J. van (2019) Forecasting bus ridership with trip planner usage data : a machine learning application.

[img] PDF
Abstract:Public transport gives much attention to environmental impact, costs and traveler satisfaction. Good short-term demand forecasting models can help improve these performance indicators. It can help prevent denied boarding and overcrowding in buses by detecting insufficient capacity beforehand. It could be used to operate more economically by decreasing the frequency or the bus size if there is overcapacity. It could help operators plan their buses during incidental occasions like big public events where little information is known and it can finally be used to reliably inform the travelers on the current crowdedness. This thesis investigates the usefulness of a new data source; the usage data of a trip planner for public transport. The data of 9292 was used. 9292 is one of the major trip planners in the Netherlands and includes all public transport modes for the whole country. A regression analysis is used to determine the forecasting potential of the trip planner usage data. This data is regressed towards smart card transaction data. 5 machine learning models were trained using different data partitions. With the models it is possible to forecast the number of passengers on a bus.
Item Type:Essay (Master)
Faculty:BMS: Behavioural, Management and Social Sciences
Subject:31 mathematics, 54 computer science, 55 traffic technology, transport technology
Programme:Industrial Engineering and Management MSc (60029)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page