PENERAPAN METODE K-NEAREST NEIGHBOR (KNN) UNTUK KLASIFIKASI CUACA DI WILAYAH TANJUNGPINANG

Abstract

The weather in Tanjungpinang area can be known by classification using previous data to study weather patterns that occur. Classification can be applied using the K-Nearest Neighbor method. K-Nearest Neighbor is a supervised learning algorithm. The purpose of this algorithm is to classify new objects obtained based on attributes and samples from training data. In this study KNN was used with parameters of temperature, humidity, pressure, and wind speed obtained from The Meteorological Station of Raja Haji Fisabilillah Tanjungpinang. Data taken from January 2016-December 2019. The result of this study is a classification consisting of weather types namely Thunderstorm, Lightning, Mist, Smoke, Cloudy, Prec in Sight, Rain, Haze, Shower, Lightning Rain. The results of this study obtained the highest accuracy of 70% of k = 31, 33, 35, 37, 39, 41, 45, 47, 50 and the lowest accuracy of 30% of k = 7