Analysis of Travel Mode Choice in Seoul Using an Interpretable Machine Learning Approach

Table 2

Descriptive statistics of the variables.

Variable name	Category	%	Mean	Standard deviation
Travel mode	Car	18.5
	Bike	2.5
	Transit	35.3
	Walking	43.7
Trip-related attributes
Activity duration (min)			490.2	251.1
Travel time (min)			21.7	15.9
Departure time	Peak	50.6
	Nonpeak	49.4
Trip type
	HBW	31.8
	HBO	16.7
	NHBO	4.8
	RH	46.7
Tour-related attributes
Sum of activity duration (min)			509.2	235.6
Sum of travel time (min)			51.6	33.4
Number of trips	2	70.9
	3	10.5
	4	16.4
	5	1.4
	6	0.8
Tour type
	HWH	51.6
	HOH	27.0
	HOWH	21.4
Individual attributes
Age			44.6	20.0
Gender	Female	51.7
	Male	48.3
Car owner	Yes	72.0
	No	28.0
Driver’s license	Yes	54.7
	No	45.3
Income	High	33.0
	Low	67.0
Built environment attribute
Land use in D: residential			0.49	0.20
Land use in D: commercial			0.29	0.20
Population density at D			42,862	11,771
Number of workers at D			32,787	75,271
Number of bus stops at D			125.2	85.5
Number of subway stops at D			1.0	1.2

Note. D = destination of a trip.