BuzzFeed News trained a computer to find spy planes by letting a machine-learning algorithm sift for planes with flight patterns that resembled those operated by the FBI and the Department of Homeland Security. Fist, they modeled the flight characteristics of almost 20,000 planes, their turning rates, speeds and altitudes flown, the areas of rectangles drawn around each flight path, and the flights’ durations. Then, they included information on the manufacturer and model of each aircraft, and the four-digit codes emitted by the planes’ transponders. Finally, they trained an algorithm called the “random forest” to distinguish between the characteristics of FBI and DHS planes and 500 randomly selected planes. Voila! The yield of all this was a calculated probability that a particular aircraft was flown by the FBI or DHS. Although there were errors in the predictions, the process proved very effective for the initial screening of possible spy planes.