Optimasi Cluster Pada K-Means Clustering Dengan Teknik Reduksi Dimensi Dataset Menggunakan Gini Index

Abstract

In K-Means Clustering, the number of attributes of a data can affect the number of iterations generated in the data grouping process. One of the solutions to overcome these problems is by using a reduction technique on the dimensions of the dataset. In this study, the authors apply the Gini Index to perform attribute reduction on the data set to reduce attributes that have no effect on the dataset before clustering with K-Means Clustering. The dataset used to be tested as a testing instrument in this research is Absenteeism at work obtained from the UCI Machine Learning Repository, with 20 attributes, 740 data records and 4 attribute classes. The results of the tests in this research indicate that the number of iterations obtained from the comparison of tests using the K-Means in a Conversional (Without Attribute Reduction) is obtained by the number of 9 iterations, while the K-Means with attribute reduction with the Gini Index obtains the number of iterations totaling 6 iterations. Clustering evaluation was calculated using Sum of Square Error (SSE). The SSE value in K-Means Clustering in a Conversional (Without Attribute Reduction) is 1391.613, while in K-Means Clustering with attribute reduction with a Gini Index, it is 440.912. From the results of the proposed method, it is able to reduce the percentage of errors and minimize the number of iterations in K-Means Clustering by reducing the dimensions of the dataset using the Gini Index