Modeling of intra-city transport choice behaviour in Budapest, Hungary

The discrete choice modeling and decisions-tree technique are used to understand the travel behavior of people in Budapest. The discrete choice modeling is applied to develop transport mode choice that mimics the travel behavior of people using their personal and travel characteristics. A revealed preference (RP) survey was conducted by the Hungarian Census Bureau which contains information about the households in Budapest in 2014, is used. Understanding the daily main trips of people is firstly analyzed using decisions-tree technique, where the impact of each variable is presented based on its importance in affecting the travel choice mode. In random utility theory, travelers choose one option out of certain available options to them to maximize their utility. The Multinomial Logit (MNL) model is used to examine the relationship between various variables connected to travelers in order to understand the travel behavior pattern. The result of the analysis shows a clear pattern between car ownership and each of family size, and age, and trip cost variables. The result of decision tree analysis demonstrates that travelers’ trip duration is the most important factor that has impact on their transport mode choice, which is mainly distributed within private cars, public transportation, and walking. The developed multinomial transport mode choice model includes sociodemographic, economic, and travel characteristics. The trip time, trip cost, age, car ownership index, trip purpose, gender, employment, and income are the main determinants the impact the transport choice mode. The developed models (i.e., decision tree and the multinomial transport logit choice) are beneficial for the decision-makers who can used it in predicting the travel demand.