Abstract

The National Aeronautics and Space Administration (NASA) is a United States state agency responsible for the space program. The state agency observes space objects including asteroids. Information from www.kaggle.com There are an infinite number of space objects, some of which are closer than we think, although we may think that 70,000 Kilometers are not dangerous to us, but on an astronomical scale it is a very small and disturbing distance. natural phenomena and including dangerous. Judging from the infinite number of objects in outer space that will cross the earth's orbit, predictions are needed to determine the level of danger or harm when crossing the earth's orbit. Prediction is an activity that can know or predict what will happen in the future which aims to find out the approximate asteroids that will cross the earth in the future. In this study, data mining classification techniques and the K-Nearest Neighbor algorithm are used to make a prediction system for the threat of asteroids when they cross the earth. Classification is the task of mining data into groups of data by classifying data items into predefined class labels, building a classification model from the data set, building a model that is used to predict future data. To determine the distance of the asteroid threat across the earth, data mining classification techniques and the K-Nearest Neighbor algorithm are used. The data that will be used is 3,500 data sourced from NASA, the data will be processed using the K-Nearest Neighbor algorithm. The results obtained are 57.71% accuracy, 54.89% precision, 81.42% recall, and 47.45% missclassification rate.

Kata kunci

Data Mining, Classification, K-Nearest Neighbor, Nasa, Asteroids, Prediction

Syahrul Firdaus

Affiliation

Conference Tracks

Refine Results

PREDICTION OF ASTEROID HAZARD DISTANCE THROUGH EARTH'S ORBIT USING K-NEIREST NEIGHBOR METHOD