Comparing decision tree algorithms to estimate intercity trip distribution
Abstract
An alternative method for estimating travelers' destination choice is proposed. Traditional methods usually do not take into account disaggregate factors. Features of the origin city along with individual factors are here considered. The use of decision tree algorithms (CHAID and CART) is presented. Both decision tree algorithms outperformed the gravity models.
- Publication:
-
Transportation Research Part C: Emerging Technologies
- Pub Date:
- April 2017
- DOI:
- 10.1016/j.trc.2017.01.009
- Bibcode:
- 2017TRPC...77...16P
- Keywords:
-
- CHAID;
- CART;
- Trip distribution modeling;
- Gravity models;
- Disaggregated analysis