Research Article

Application of Stochastic Gradient Boosting Approach to Early Prediction of Safety Accidents at Construction Site

Table 2

Factors of construction accidents on-site.

VariableTypeFeature

Dependent variable (output)Type of accidentNominalFall (1), wipeout (2), hit (3), cut (4), narrowness (5), crash (6)
Independent variable (input)GenderNominalMale (1), female (2)
Day of the weekWeekdays (1), weekend (2)
LocationSeoul (1), capital area (except for Seoul) (2), etc. (3)
NationalityKorean (1), foreigner (2)
SeasonSpring (1), summer (2),
Autumn (3), winter (4)
Time06–12 hour (1), 12–18 hour (2), 18–22 hour (3), 22–06 hour (4)
Type of employmentRegular (1), nonregular (2)
Type of occupationManagement (1), engineering (2), labor (3), operation (4)
Type of projectResidential and commercial (1), public and healthy (2), etc. (3)
AgeNumericalInteger (17–83)
Project durationInteger (1–13,752)
Project scaleInteger (0.1–2,000)
Rate of processReal number (5–90)
WorkforceInteger (5–3,000)
Year of serviceReal number (0.8–30)