Kombinasi Pembobotan Symmetrical Uncertainty Pada K-Means Clustering Dalam Peningkatan Kinerja Pengelompokan Data

Suranta Bill Fatric Ginting (Universitas Sumatera Utara, Medan)
Sawaluddin Sawaluddin (Universitas Sumatera Utara, Medan)
Muhammad Zarlis (Universitas Sumatera Utara, Medan)

JURNAL MEDIA INFORMATIKA BUDIDARMA
Vol 6, No 1 (2022): Januari 2022

Based on several studies that examine the K-Means Clustering method, it was found that in K-Means Clustering one of the weaknesses lies in the process of determining the center point of the cluster which also has implications for distance calculations in determining the similarity between data to obtain conclusions from the data. a cluster. And this is also caused by the influence of the percentage of the attributes used. If the attributes used are less relevant to their level of influence and also have a low contribution to the data, this can have a significant impact on the results of clustering. So from these problems, in this research, the author proposes to use the method in calculating the weight of data attributes in the clustering process, namely using Symmetrical Uncertainty. To test the proposed method, this research uses a dataset from UCI Machine Learning which consists of Iris with 150 data and Wine Quality with 178 data. The evaluation of the proposed clustering performance is based on the Davies-Bouldin Index (DBI) value. The test results in this study show that the proposed method can produce a significantly smaller Davies-Bouldin Index (DBI) value.