This article reports on a study that demonstrated a use of the Highway Safety Information System (HSIS) database to determine safety criteria for selecting a smart corridor using random forest, a machine-learning approach. HSIS contains a rich dataset, which includes various variables from many aspects of transportation. In this study, the authors implemented the random forest algorithm to finalize 13 safety criteria for selecting a smart corridor out of 111 variables in the HSIS and the Highway Performance Monitoring System (HPMS) from Washington State.
Guo, X., Peng, Y., & Ma, C. (2020). Safety Criteria for Selecting a Smart Corridor: Random Forest Approach Using HSIS Data from Washington State. ITE Journal, 90(12), pp 35-44.