Factor investigation of serious accidents in traffic accident data

Abstract

Although the number of traffic accidents is decreasing year by year owing to the safety per- formance improvement of automobiles, traffic accidents are still a major problem in our day-to-day life. In this study, aiming to prevent serious accidents, we analyze the collection of traffic accidents data by several data mining techniques for extracting potent factors and features of serious accidents. The real traffic accident data are preprocessed to be useful for the data analysis. The potent factors and features of serious accidents are investigated in the data by fuzzy reasoning models, SVM, random forest, and rough set analysis.