A Random Forest Algorithm-based Accident Prediction to Prevent Marine Pilot Occupational Accidents

  • Gokhan Camliyurt (Department of Navigation, Graduate School, Korea Maritime, and Ocean University) ;
  • Won Sik Kang (Transportation Safety Assessment Office, Korea Maritime Transportation Safety Authority) ;
  • Daewon Kim (Shipping and Logistics Research Division, Korea Maritime Institute) ;
  • Sangwon Park (Division of Navigation Convergence Studies, Korea Maritime, and Ocean University) ;
  • Youngsoo Park (Division of Navigation Convergence Studies, Korea Maritime, and Ocean University)
  • 발행 : 2022.06.02

초록

Marine pilot occupational accidents during transfer to/from the ship are at the top of the agenda after several safety campaigns by IMPA and individual attemptsThere is multiple transfer method for the marine pilot, but a most common way is to use the pilot cutter. This paper aims to predict marine pilot occupational accidents before it occurs by using historical data. Since the problem depends on several variables, this paper develops a model by using the random forest method to predict marine pilot accidents before happening with the random forest method by using RStudio software

키워드