Identifying risk factors associated with the pipeline failure using machine learning methods

Abstract
In this study, we employ two machine learning models namely Neural Networks and Random Forest to examine the impact of geographical and meteorological factors on the pipe failure. We find that the most important factors related to pipeline failure are snow on ground, total snow, and total rain, where among them snow on ground seems to be the most important variable in both models; and geographical factors do not contribute significantly to the pipeline failure.