Travel Mode Choice Analysis Based on Household Mobility Survey Data in Milan: Comparison of the Multinomial Logit Model and Random Forest Approach

The identification of household travel mode choice is important for travel demand management/mobility management. According to the household travel survey data collected in Milan Municipality in 2005-2006, this paper investigates the household travel mode choice characteristics among public transport (PT), private car, usage of PT and private car simultaneously and other, and further, analyzes the factors determining travel mode choice including household size, vehicle ownership, household income, travel distance and travel time. The analysis is based on two approaches: the multinomial logit model (MNL) and the random forest (RF) approach. Under the full sample size, the precision reaches 71.75% with the MNL model to estimate the household travel mode choice, which is in contrast with the precision of 73.48% with the RF approach. To estimate the household travel mode choice, it is found that the precisions of the two approaches fluctuate obviously with the size of sample, and surprisingly, the increased sample size will decrease the precisions. The precisions of thirty-one MNL models considering different combinations of impact factors are investigated, and the importance ranking of impact factors are identified. MNL model demonstrates that vehicle ownership is the most critical factor influencing household travel mode choice, followed by travel distance, travel time, household size and household income, which is consistent with the RF approach by using Random Forest package in R software directly.

• Supplemental Notes:
  • This paper was sponsored by TRB committee ADB40 Standing Committee on Transportation Demand Forecasting.
• Authors:
  • Liang, LeiLei
  • Xu, Meng
  • Grant-Muller, Susan
  • Mussone, Lorenzo
• Conference:
  • Transportation Research Board 97th Annual Meeting
  • Location: Washington DC, United States
  • Date: 2018-1-7 to 2018-1-11
• Date: 2018

Language

• English

Media Info

• Media Type: Digital/other
• Features: Figures; References; Tables;
• Pagination: 8p

Subject/Index Terms

• TRT Terms: Automobile travel; Mobility; Mode choice; Multinomial logits; Public transit; Travel surveys
• Geographic Terms: Milan (Italy)
• Subject Areas: Highways; Planning and Forecasting; Public Transportation;

Filing Info

• Accession Number: 01658371
• Record Type: Publication
• Report/Paper Numbers: 18-03595
• Files: TRIS, TRB, ATRI
• Created Date: Jan 29 2018 10:27AM