Perbandingan Analisis Cluster K-Means dengan Metode Ward dalam Pengklasteran Cuaca di Stasiun Pengamatan BMKG

Abstract

Weather cluster analysis is an application of cluster analysis on weather data to group similar weather patterns. It is conducted to understand the existing weather patterns in historical or current data. This research was carried out at the Central Bureau of Statistics of North Sumatra Province, Jl. Asrama No.179, Dwi Kora, Kec. Medan Helvetia, Kota Medan, North Sumatra 20123. The research was conducted over a period of 1 month. The standard deviation ratio values in KMeans and Ward methods of Cluster Analysis yielded the same result, which is 0.137. Based on the standard deviation ratio values, it appears that both methods used are equally effective. Therefore, according to this research, both methods are highly recommended for use in weather’s data.