We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Pendekatan Initial Centroid Search Untuk Meningkatkan Efisiensi Iterasi Klustering K-Means.
- Authors
Nasution, Muhammad Zulfahmi; Hasibuan, Muhammad Siddik
- Abstract
K-means clustering aims to gather a set of optimal initial centroid through successive iterations. The fact that more optimal the point of the initial cluster centers, the less iteration of the K-Means clustering algorithm will be needed for convergence. Therefore, one way to find a better initial set of centroid is through an iterative approach to searching for a better set of initial centroid for K-Means clustering. The first step we will take is to take data samples from the data set and run the short runs of the K-Means clustering algorithm on it (not for convergence) but as the initial process of centroids initialization. Then we will repeat the short runs as an iteration process with a number of initial centroid cluster being randomly initialized before and measuring within-cluster by sum-of-squares-error to determine the goodness of cluster membership. The final centroids that provides the lowest inertia will continue to complete the k-means grouping process. Our hope is that this approach will lead to a better initial centroid set for the k-means clustering to improve the performance of the K-Means Algorithm because the convergence results of the K-Means Algorithm will be directly proportional to the selection of initial centroids.
- Subjects
K-means clustering; SUM of squares; CENTROID; GROUP process
- Publication
Techno.com, 2020, Vol 19, Issue 4, p341
- ISSN
1412-2693
- Publication type
Article
- DOI
10.33633/tc.v19i4.3875