Using Mobility Profiles for Synthetic Population Generation

Bektas, Alperen; Schumann, René (22 September 2019). Using Mobility Profiles for Synthetic Population Generation In: Social Simulation Conference (SSC). Mainz Germany. 23-27. September 2019.

Full text not available from this repository. (Request a copy)

Agent-based modeling (ABM) is a wide-spread technique that can be utilized as an artificial laboratory for in-silico experiments of real-case studies of different domains such as mobility. To initialize agent/environment attributes and their relationships, disaggregated (individual level) micro-data is required as an input. However, having such data is not often possible due to several reasons such as privacy concerns. To bridge the gap, generating realistic synthetic data (from census/survey data) becomes an initial and essential step of agent-based modeling. In this piece of research, we employ the mobility profiles of the Swiss population for generating synthetic populations along with their mobility activities. To validate the synthetic data, an agent-based model, which is already calibrated to the empirical data, is re-run with a sample and the generated synthetic data. Accumulated decisions of agents in both cases are compared. In addition, marginal frequencies of control attributes are benchmarked. The first obtained results demonstrate that increasing the size of the generated population decreases the difference between simulation results of the synthesized data and the real data.

Item Type:

Conference or Workshop Item (Paper)

Division/Institute:

Business > Institute for Public Sector Transformation > Daten und Infrastruktur

Name:

Bektas, Alperen and
Schumann, René

Subjects:

Q Science > QA Mathematics > QA75 Electronic computers. Computer science

Language:

English

Submitter:

Alperen Bektas

Date Deposited:

23 Dec 2020 07:06

Last Modified:

23 Dec 2020 07:06

Uncontrolled Keywords:

Synthetic population generation·Agent-based modeling·Mobility profiles·Cluster analysis·Demand modeling

URI:

https://arbor.bfh.ch/id/eprint/13915

Actions (login required)

View Item View Item
Provide Feedback