Prediction Of Asteroid Hazard Distance Through Earth's Orbit Using K-Neirest Neighbor Method

Abstract

The National Aeronautics and Space Administration (NASA) is the U.S. government agency that is responsible forspace program. NASA observes objects in space, including asteroids. Asteroids are small, rocky objects that orbit thesun with irregular shapes and are also called planetoids. The Government agencies observe space objects includingasteroids. In terms of the infinite number of objects in space that will cross Earth's orbit, prediction is needed todetermine the danger and its level when they are crossing Earth's orbit. Prediction is a process to know what willhappen in the future which is aimed to find out the approximate asteroids that will cross the earth in the future. In thisstudy, data mining classification techniques and the K-Nearest Neighbor algorithm are used to create a predictionsystem for the threat of asteroids while crossing the earth. Classification is a grouping by classifying items intodesignated class labels, building a classification model from the data set, building a model that is used to predict futuredata. To determine the distance of the asteroid's threat throughout the earth, data mining classification techniques andthe K-Nearest Neighbor algorithm are used. The results are 57.71% accuracy, 54.89% precision, 81.42% recall, and47.45% missclassification rate.