Analysis of Centroid Value Variations Against the Number of Iterations Using the Clustering K-Means Algorithm
H. Simanjuntak1, M Zarlis2, P H Putra3
1H. Simanjuntak, Student, Department of Information Technology, Faculty of Computer Science and Information Technology, Universitas Sumatera Utara, Indonesia.
2P H Putra, Student, Department of Information Technology, Faculty of Computer Science and Information Technology, Universitas Sumatera Utara, Indonesia.
3M Zarlis, Department of Information Technology, Faculty of Computer Science and Information Technology, Universitas Sumatera Utara, Indonesia.
Manuscript received on 09 May 2019 | Revised Manuscript received on 19 May 2019 | Manuscript Published on 23 May 2019 | PP: 1356-1358 | Volume-7 Issue-6S5 April 2019 | Retrieval Number: F12350476S/2019©BEIESP
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: In this study, the researchers used the K-means Algorithm to see and examine the effect of variations in centroid values on the number of iterations. The results of this study were: Different number of centroid results in different number of iterations. The large number of centroid values did not always cause the number of iterations to increase. Testing with the number of centroid values 3, 8, 10 has a smaller number of iterations, namely the 3 iteration with the level of similarity of the previous data, compared with the number of centroid values 2, 3, 4, 5, 6, 7, 8, 9, 10. Testing with the number of centroid 2 values had a greater number of iterations, reaching the 9th iteration to reach the previous level of data similarity.
Keywords: Algorithm Clustering Analysis Iterations Data.
Scope of the Article: Algorithm Engineering