This morning I was listening to a podcast from Marketplace from a couple of days ago and heard an interesting piece on using machine learning to identify where lead pipes might be located in the beleaguered Flint, MI. Alexis Madrigal provided a follow-up in the Atlantic to a proceeding paper in which the original authors described a technique for identifying, with 80 % accuracy, lead pipe locations in Flint using different “machine learning” techniques – specifically, a form of…